An integrated value-risk investigation of contactless mobile payments adoption

Accepted Manuscript An Integrated Value-Risk Investigation of Contactless Mobile Payments Adoption Mihail Cocosila, Houda Trabelsi PII: DOI: Reference:

S1567-4223(16)30073-4 http://dx.doi.org/10.1016/j.elerap.2016.10.006 ELERAP 691

To appear in:

Electronic Commerce Research and Applications

Received Date: Revised Date: Accepted Date:

12 January 2016 26 October 2016 29 October 2016

Please cite this article as: M. Cocosila, H. Trabelsi, An Integrated Value-Risk Investigation of Contactless Mobile Payments Adoption, Electronic Commerce Research and Applications (2016), doi: http://dx.doi.org/10.1016/ j.elerap.2016.10.006

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Manuscript title: An Integrated Value-Risk Investigation of Contactless Mobile Payments Adoption Authors: Mihail Cocosila, PhD – corresponding author Associate Professor Faculty of Business, Athabasca University, 1 University Drive Athabasca, Alberta, T9S 3A3 Canada 1-780-675-6189 [email protected] http://business.athabascau.ca/faculty/mihail-cocosila-phd/ Houda Trabelsi, MSc Academic Coordinator Faculty of Business, Athabasca University, 1 University Drive Athabasca, Alberta, T9S 3A3 Canada 1-780-675-6189 [email protected] http://business.athabascau.ca/faculty/houda-trabelsi-msc/ Short Bios: Dr. Mihail Cocosila is an Associate Professor at the Faculty of Business, Athabasca University, Canada. His research interests include information technology adoption, human-computer interaction, and mobile and wireless services. He published in Electronic Markets, Communications of the ACM, Communications of the AIS, and Canadian Journal of Administrative Sciences. Ms. Houda Trabelsi is an Academic Coordinator at the Faculty of Business, Athabasca University, Canada. Her research interests are in management information systems and associated topics.

Abstract 1

The objective of this study is to investigate empirically consumer adoption views on credit card contactless payments with smartphones. Contactless Near Field Communication (NFC) mobile payments have unquestionable advantages but may also face some user doubts. To investigate consumer possible dual views about this new mobile service, a third-order innovative factor integrating value and risk perceptions was developed and validated empirically in an adoption model through a survey involving 289 participants. Findings indicate that the integrated valuerisk perception is a significant factor of adoption of NFC payments with smartphones having utilitarian and enjoyment values as the main user motivators, and psychological and privacy risks as the most important deterrents. Overall, this study proposes a broader view of the consumer perceptions of value by integrating value and risk into an unbiased model of technology adoption. Keywords Near field communication mobile payments, Radio Frequency Identification, Perceived value, Perceived risk, Adoption

2

An integrated value-risk investigation of contactless mobile payments adoption

Abstract The objective of this study is to investigate empirically consumer adoption views on credit card contactless payments with smartphones. Contactless Near Field Communication (NFC) mobile payments have unquestionable advantages but may also face some user doubts. To investigate consumer possible dual views about this new mobile service, a third-order innovative factor integrating value and risk perceptions was developed and validated empirically in an adoption model through a survey involving 289 participants. Findings indicate that the integrated value-risk perception is a significant factor of adoption of NFC payments with smartphones having utilitarian and enjoyment values as the main user motivators, and psychological and privacy risks as the most important deterrents. Overall, this study proposes a broader view of the consumer perceptions of value by integrating value and risk into an unbiased model of technology adoption.

Keywords Near field communication mobile payments, Radio Frequency Identification, Perceived value, Perceived risk, Adoption

1. Introduction Contactless credit card payments with smartphones using Radio Frequency Identification (RFID) and Near Field Communication (NFC) technologies are set to become the most common means of electronic payment in mobile commerce transactions in the near future (Rose 2012). NFC-based mobile payments using smartphones is, reportedly, going to increase from $4 billion in 3

2012 to $191 billion in 2017, breaking the $100 billion mark in 2016 globally (ABI 2015). Unsurprisingly, this happens in a world seeing a continuous increase of mobile phone popularity – thus the global number of mobile phone users will be reaching 9.3 billion by the end of 2018, hence surpassing the entire world population (Li et al. 2014). For instance, market analysis companies forecast high figures for Canada alone on the use of RFID-equipped smartphones for contactless credit card payments: about 80% of the smartphones in use will be NFC-enabled while the total of such payments would reach $14.2 billion by 2016 (Canada Newswire 2015). Contactless mobile payments are, thus, anticipated to increase in popularity as they are envisioned as a convenient, open, safe and secure system supported by clear standards and operating frameworks (Canada Newswire 2015). Facilitating conditions created as a result of joint efforts of mobile phone industry, wireless providers and credit card companies are expected to amplify user choices and satisfaction and, consequently, accelerate the adoption of NFC-based mobile payments that look as a beneficial service for all stakeholders. However, it is important to have a more comprehensive understanding of the implications of contactless mobile payments beyond this optimistic picture by also looking at possible barriers to user acceptance as credit card use itself is sometimes feared as posing various risks and threats. Therefore, a new Information and Communication Technology (ICT) application based on a more sophisticated use of credit cards might exacerbate some of the already negative user perceptions. In general, it is well-established in Information Systems (IS) research that, beyond technology and business aspects, user views on an ICT are a key determinant for the eventual success of that technology (Venkatesh et al. 2002). Furthermore, a service innovation could be successful only when it is accepted by the market (Burgelman et al. 2004). Research has shown that there are many factors that could affect the perceptions and attitude of target consumers on a new ICT and 4

ensure its acceptance and ultimate success. While positive views like the perception of usefulness or of ease of use are favorable factors, negative views like the perception of risk or lack of control are deterrents for the use of a new ICT solution. Accordingly, a comprehensive adoption investigation should integrate both favoring and disfavoring factors in a theoretically sound research model. As no scientific investigation regarding user perceptions of the incipient NFC-based mobile payments was found in the literature available to date, the objective of this study is to conduct an empirical research to identify the most important positive and negative factors in the adoption equation. For that, a theoretical model including an innovative value-risk factor and contrasting the perceived gains to the costs of using NFC mobile payments is built. The model is then validated through a quantitative empirical research that involves surveying online 289 Canadians. This article first describes the NFC payments technology approach, then presents a theoretical background on perceived value and perceived risk, and then develops an original ICT adoption model relying on an integrated value-risk construct. Finally, the study methodology, main findings and discussion and conclusions are presented.

2. How contactless NFC payments with smartphones work? Contactless payments with smartphones combine the RFID technology with the mobile communication technology (De Reuver et al. 2015, Lee et al. 2015). Following the incorporation of the RFID contactless credit card standards into the NFC standard, users are able to pay for their purchases by holding an NFC smartphone next to contactless readers in various places such as stores or gas stations as an alternative to using their ‘tap and pay’ credit cards (NFC World Forum 2015).

5

To use contactless mobile payment technology, consumers need to have a contactless RFID credit card from their bank, acquire a NFC-enabled smartphone, and download a mobile payment application on their phone. An NFC-enabled smartphone has a dedicated chip and an antenna for radio communication within a short distance. Information about the owner’s contactless credit cards will be encrypted and stored in a secure area in the phone’s memory. When making a payment the consumer taps their smartphone on the contactless payment-capable point of sale (POS) system, similarly to using a contactless credit card. The smartphone communicates through NFC technology with the merchant’s POS and transmits the information necessary to complete the payment. Due to their convenience, contactless mobile payments have a huge market potential and are already attracting the attention of telecom operators, banking institutions and ICT service providers worldwide (e.g., Japan, South Korea, Kenya, Europe and the US) (Zhu and Chen 2011). In Canada, for instance, the Toronto-based lender Canadian Imperial Bank of Commerce (CIBC), with international operations in the US, the Caribbean, the UK, and Asia, introduced the first NFC application for mobile payments for Visa and MasterCard in 2012 jointly with the cell phone manufacturer BlackBerry and the telecom operator Rogers Communications (Canada Newswire 2015, Visa Pay Wave 2015). Contactless smartphone payment is hoped to become very popular in just a few years’ time due to the envisioned advantages for all stakeholders, starting with consumers. However, a technology combining the benefits of credit card payments and the enjoyment of smartphone use may also exacerbate consumer potentially negative views regarding the two parties - e.g., fears of privacy threats or anxiety of using an unknown mobile banking service. These warrant the investigation of user perceptions on contactless mobile payments through an unbiased theoretical approach considering both key positive and negative factors relative to user adoption. 6

3. Theoretical background Investigating user adoption of a new ICT has constantly been a fundamental area of research in information systems with several theories and models being proposed and successfully validated over time (Venkatesh et al. 2002), including for the domain of mobile payments research (Dahlberg et al. 2015). Theories and models investigated have usually been taking into consideration a number of user factors, basically all being favorable antecedents of the intention to adopt a new ICT. Thus, in most of the cases, a technology would be adopted by users if they expect a high performance and low effort associated with its use, if people see a positive influence toward that use from significant other persons as well as general conditions facilitating the use (Venkatesh et al. 2003). In recent years, as new technologies and ICT services have been offered in increasing numbers to the general population, some IS researchers have considered value-based models as an approach to investigating ICT adoption. These researchers believe that the ‘traditional’ adoption models, like the Technology Acceptance Model (TAM) proposed by Davis (1989), have limited ability or flexibility in explaining adoption of new ICT by general consumers (Jung 2014). Thus, in contrast to technology users in organizations looking mostly at improving their work performance with the help of an easy to use technology, consumers would rather be concerned about the net value of the technology for them (Kim et al. 2007, Lin et al. 2012). Consequently, adoption intention regarding a new ICT could be more appropriately explained by assessing the value consumers perceive in using that technology. For instance, compatibility with consumer values was identified as an important factor favoring e-banking adoption in a systematic analysis of 247 peer-reviewed articles published in three decades of research on this topic (Hoehle et al. 2012).

7

The concept of perceived value is popular in consumer behavior studies and expresses a tradeoff measure between the gain brought by a product or activity and the cost (or “sacrifice”) of these (Zeithaml 1988). Consumer assessment of value is an important predictor of the intention to purchase – i.e., the higher the value perceived in a good or service (i.e., users see more benefits than costs (Hsu and Lin 2015)) the higher the likelihood to make the respective purchase (Kim et al. 2011, Park and Lee 2011, Sheth et al. 1991, Zeithaml 1988). Following the same line of thinking, perceived value in an ICT adoption research contrasts the factors favoring adoption (as gain) and the factors disfavoring adoption (as cost). Consequently, the higher the value perceived the higher the probability to adopt various new ICTs such as mobile devices (Jung 2014, Yang et al. 2016) or services (Al-Debei and Al-Lozi 2014, Hsu and Lin 2015, Kuo et al. 2009, Turel et al. 2010, Xu et al. 2015). Perceived value in consumer behavior usually contrasts the gain brought by using a product or service to its cost expressed in monetary units (i.e., what the price of that product or service is). However, according to previous research, a more refined view of the cost should consider that there may be both a perceived monetary sacrifice (e.g., financial effort) and a perceived nonmonetary sacrifice (e.g., time, physical or mental effort) that determine the ‘give’ side of the value (Lin et al. 2012, Xu et al. 2015, Zeithaml 1988). Following this perspective, and consistent with some of the previous research in consumer behavior (Snoj et al. 2004, Sweeney et al. 1999), an innovative view of the perceived value of an ICT service may consider including perceived risk on the cost side in order to capture a broader perception of the consumer sacrifice to adopt that service, beyond the purely monetary aspect. This approach may be particularly suitable for newer ICT services where there may be some negative user perceptions that do not translate into immediate monetary consequences. This may be the situation of the adoption of contactless NFC smartphone payments that do not involve a fee but, if 8

appropriate measures are not taken, may have negative consequences in other directions (e.g., security and privacy) seen as perceived risks. In particular, risk was identified by Hoehle et al. (2012) as a very important obstacle to e-banking adoption in a systematic analysis of publications in the last three decades. The negative influence of this factor on consumer intentions is further confirmed by a meta-analysis of mobile banking adoption literature reported by Shaikh and Karjaluoto (2015). Perceived risk sources from consumer behavior research and expresses a potential loss as seen by individuals, hence not necessarily real (Lim 2003), when seeking a purchase. It became popular in IS research as well in recent years, especially in technology adoption studies (Featherman and Pavlou 2003) together with resistance to adoption (Lapointe and Rivard 2005). Thus, as new ICT artifacts have become more and more refined and complex, researchers have also detected the influence of some risk factors disfavoring the adoption by capturing user various fears (e.g., of wasting time or money or of being exposed to physical or social discomfort with that ICT), even if these fears may not correspond to actual dangers. Previous research demonstrated that such opposing factors cannot be ignored in a more complex investigation of the adoption of a new technology. Thus, perceived risk was usually appended as an antecedent to the constructs in popular adoption models like the Technology Acceptance Model or the Motivational Model and was proven to have a negative total effect on the intention to adopt a new ICT (Cocosila et al. 2009, Featherman and Pavlou 2003, Kim and Han 2008). A more refined step into accounting for the negative factors of the technology adoption equation for consumers would be to consider these factors in an unbiased theoretical model at the same level of importance with that of the positive determinants. This could be achieved by integrating perceived risk of adopting a new ICT (like NFC mobile payments) into the value

9

perception in order to obtain a comprehensive model of consumer adoption views. Therefore, this study proposes the following research question: What is the combined effect of perceived user value and risk on the adoption intention of contactless smartphone payments?

4. Research model and hypotheses development This research proposes to integrate perceived risk into perceived value in order to investigate consumer views on contactless payments with NFC-enabled smartphones. Stemming from the concept of perceived value in consumer behavior research, as discussed in Theoretical background section above, perceived value of an ICT service should have two sides - the gain brought by the service use and the cost associated with that. The theoretical model, having as key component an innovative construct capturing the integrated value-risk perception, and the hypotheses proposed in this research are depicted in Figure 1.

10

Fig. 1. Theoretical model and hypotheses

Previous research in consumer behavior acknowledged value as a comprehensive multisided perception combining cognitive and emotive aspects of individuals’ views relative to consumption of goods or services in various choice situations (Kim et al. 2011, Rezaei and Ghodsi 2014, Sweeney and Soutar 2001). It has also been well-established that value is seen as a difference between the gain and the cost relative to a product or activity (Zeithaml 1988). Drawing from this foundation, in IS research the gain side of the value associated with the use of an ICT product or service can be expressed broadly as a combination of an extrinsic and cognitive benefit on one hand and of an intrinsic and affective benefit on the other hand (Kim et al. 2007). While the extrinsic and cognitive benefit is related to the perception of the utilitarian value of a product or

11

service (i.e., reaching some external goals through its use), the intrinsic and affective benefit is an expression of the hedonic or emotional value of that good or service (Al-Debei and Al-Lozi 2014, Kim and Han 2009). This broad view of the benefit side of the value rhymes with the two key facets of motivation (i.e., extrinsic and intrinsic) in the Motivational Model also used in IS research to explain the adoption of a new technology (Davis et al. 1992, Xu et al. 2015). In addition to these, a third benefit that may be significant for user views on an ICT regards social aspects of the value. Thus, consumer behavior and information systems literature show that, in addition to utilitarian and emotional sides, there is a social dimension of the value that captures the improvement of the image of self associated with the use of a product or service (Sheth et al. 1991, Sweeney et al. 1999) as purchasing these goods has also a meaning in the buyer’s community. For instance, purchasing a certain product or service (like a new mobile phone application) may increase the economic and social status, and, consequently, self-concept (Kuo et al. 2009, Sweeney and Soutar 2001). Accordingly, it is hypothesized that the gain side of the model integrating value and risk has three key factors, i.e.: H1: The utilitarian component of the value will have a positive effect on the integrated value-risk perception. H2: The enjoyment component of the value will have a positive effect on the integrated value-risk perception. H3: The social component of the value will have a positive effect on the integrated valuerisk perception. Researchers agree that perceived value is a multidimensional formative factor that could be influenced by many benefit and sacrifice variables (Lin et al. 2005, Snoj et al. 2004). An underexplored approach of investigating value is to use perceived risk as a broader concept for the ‘give’ (i.e., sacrifice) side of the value, instead of cost that has a more limited and predominantly 12

financial meaning. Previous research studying the influence of risk perception in value-based models considered risk as an antecedent of the value (Agarwal and Teas 2001, Agarwal and Teas 2004, Forsythe and Shi 2003, Snoj et al. 2004, Sweeney et al. 1999, Yang et al. 2016) or as a moderator of the value’s influence on the purchase intention (Chang and Tseng 2013, Chiu et al. 2014). These studies usually included limited aspects of the risk concept (e.g., predominantly financial) and reported mixed results (e.g., risk negatively affects customers’ perceived value or, conversely, there is no significant influence). In contrast to previous research where value and risk are separate constructs, and as a significant innovative theoretical approach, the current study includes a multisided perceived risk as the ‘sacrifice’ component intrinsic to perceived value. The resulting factor called “integrated value-risk perception” is believed to better capture the multidimensionality of both value and risk into an unbiased pros-cons construct that would be suitable to investigating consumer views on a new ICT having both potential advantages and disadvantages from a user perspective. Indeed, perceived risk construct has been used in consumer behavior research to express the real or virtual negative consequences seen in several directions by consumers in association with the purchase of a product or service (Laroche et al. 2003, Lim 2003, Stone and Grønhaug 1993): performance (the product or service may not work properly), financial (the purchase may be a waste of money), time (the purchase may be time consuming), health (the product or service may pose physical risk), social (significant other people for the buyer may disapprove the purchase), and psychological (buyer has general anxiety on the worthiness of the purchase). When investigating user views regarding contactless payments with smartphones, theoretical reasoning indicates that only time, social and psychological sides of the risk would worth consideration: consumers may fear wasting time if subscribing for the service, face disapproval from their family or friends when subscribing for an unknown service, and have a general feeling 13

of anxiety regarding the decision of subscribing for the NFC-enabled mobile payments. As this service is in an incipient stage, there is not enough rationale to assess performance risk. Financial risk and health risk are not concerns as the service does not involve a fee and does not pose physical risk for the users. On the other hand, an additional type of risk often reported in IS studies is expected to have a significant meaning for this research - perceived privacy risk. User concerns about the security of their personal and confidential data and fears about these data becoming available to third parties, including wrongdoers, have been traditionally highly relevant for mobile banking (Bagadia and Bansal, 2016) as well as for e-services in general (Featherman and Pavlou 2003, Grassie 2007, Yang et al. 2015). It is, therefore, expected that privacy risks play an important role in the adoption of contactless payments with NFC smartphones too (Yu et al. 2012). Following the example of previous studies using perceived risk in technology adoption research (Featherman and Pavlou 2003), the model built in this study captures risk perception as a second order overall risk construct encompassing the four meaningful facets: time, social, psychological, and privacy. Therefore, it is hypothesized that: H4a: Perceived time risk will have a positive effect on perceived overall risk. H4b: Perceived social risk will have a positive effect on perceived overall risk. H4c: Perceived psychological risk will have a positive effect on perceived overall risk. H4c: Perceived privacy risk will have a positive effect on perceived overall risk. From the understanding of the value itself it is logical that the larger the sacrifice side the lower the overall value (Zeithaml 1988). Consumers are conscious that riskier products may involve losses so they would see less value in such products (Beneke et al. 2013). This is supported by previous research showing that a higher risk for a purchase would negatively affect

14

the perceived value and, consequently, the purchase decision (Kim et al. 2008, Snoj et al. 2004, Sweeney et al. 1999). Accordingly, for the current study perceived risk would have a negative influence on the value-risk factor. Since perceived overall risk is included as a second-order construct, as mentioned previously, the resulting integrated value-risk perception is a third-order construct. Hence, it is proposed that: H5: Perceived overall risk will have a negative effect on the integrated value-risk perception. The perception of value is context specific and is believed to drive peoples’ attitudes and behaviors, thus being aligned with behavioral theories such as Theory of Reasoned Action (Fishbein and Ajzen 1975, Lee et al. 2014). Therefore, theoretical reasoning indicates that the higher the value compared to risk consumers perceive in an ICT product or service (hence there is more gain than cost) the more inclined individuals are to adopt that ICT. This relationship is also strongly supported by empirical evidence (Kim et al. 2007, Turel et al. 2007, Turel et al. 2010). Consequently, it appears logical to hypothesize that: H6: The integrated value-risk perception will have a positive effect on adoption intention.

5. Methodology A quantitative empirical study was conducted to validate the theoretical model proposed by this research. A country-wide survey was conducted across Canada and the resulting data were analyzed with Partial Least Squares (PLS) modeling. This Structural Equation Modeling (SEM) approach was selected as the main tool of data analysis following methodological discussions and recommendations mentioned in recent prominent IS literature (Gefen et al. 2011, Goodhue et al. 2012a, Goodhue et al. 2012b, Ou et al. 2014, Ringle et al. 2012). Thus, variance-based PLS was 15

preferred to covariance-based SEM tools (e.g., AMOS-LISREL approaches) because the former tool is more suited for exploratory work (Bontis et al. 2002, Ringle et al. 2012) and theory development (Chwelos et al. 2001) while the latter way is more appropriate for confirmatory testing of an existing theoretical model through field data (Chwelos et al. 2001, Ou et al. 2014). This study is considered to be predominantly exploratory and to focus on an innovative theoretical approach by integrating perceived risk as a broader concept for the ‘give’ (i.e., negative) side of the value perception, instead of the more traditional and limited financial cost. Furthermore, PLS was shown to work well with formative constructs (Thomas et al. 2005) by easily addressing their statistical problems such as identification and convergence (Chin 1998a, Ou et al. 2014, Petter et al. 2007, Ringle et al. 2012) compared to covariance-based SEM tools that imply stricter assumptions (Chwelos et al. 2001, Ou et al. 2014). The current study includes a perceived overall risk as a second order construct and an integrated value-risk perception as a third-order construct and PLS-SEM was, therefore, considered more appropriate to model this complex formative structure. These two formative constructs were measured through the repeated indicators procedure by using the indicators of the first order constructs they comprise (Lohmoller 1989). The number of participants necessary for the study resulted from an initial assessment related to the theoretical model. Thus, according to literature, the sample size necessary for a PLS-SEM analysis should be at least ten times the larger of either the number of indicators of the most complex formative construct or the number of paths leading to the endogenous construct with the largest number of antecedents in the model (Bontis 1998, Chin 1998b, Jarvenpaa et al. 2004). For the theoretical model in this research, it appears obvious that the sample size is determined by perceived value-risk that, as a third order construct, has 23 items. Accordingly, the sample size should be at least 230 valid responses. Another approach to estimate the required sample size may 16

be to use Daniel Soper’s a-priori sample size calculator for Structural Equation Models (Soper 2015). For an anticipated effect size of 0.3 (medium), a statistical power of 0.8, and a probability level of 0.05, a recommended sample size of 183 participants results for this model. Hence, from the two approaches above, a conservative estimation of the sample size requires at least 230 valid cases. Based on these, as a contingency measure for ensuring collected data accuracy, the experiment involved 300 initial respondents. These were recruited on a first come-first accepted basis from several tens of thousands of consumers pre-registered with a market survey company in Canada. This approach offered advantages in terms of ensuring appropriate coverage of the target population and increasing respondents’ willingness to participate while maintaining a reasonable cost of the data collection. The data collection process stopped when exactly 300 complete answers were registered, so the nonresponse percent and their influence could not be assessed. Including conditions asked participants to be at least 18 years old, be smartphone owners, and credit card users, hence to be adult consumers. Participants informed about the study conditions and consenting to participate were first asked to read details on contactless payments with NFC smartphones from a reliable source. Thus, half of the sample was directed to the web site of a major Canadian bank that was presenting information on mobile and contactless payments in a generally positive light whereas the other half of the sample was directed to the web site of an international media company displaying a rather cautionary view on the service. Participants were allocated randomly to the two half samples and the recruitment stopped when exactly 150 complete answers were recorded for each subsample. The two subsamples will be called ‘Pros’ and ‘Cons’, respectively, in the remaining of this study.

17

After reading the contradictory information on contactless payments with NFC smartphones, all participants were asked to complete the same online survey eliciting their views on this new mobile ICT service. The survey comprised both closed-ended questions and open-ended questions in order to deepen the understanding of the consumer perceptions. This article analyzes only the data collected through closed-ended questions related to the theoretical model. These questions targeted participant demographics as well as the items of the first-order latent variables in the research model. Consistent with previous research examining the role of perceived risk in ICT adoption (Featherman and Pavlou 2003), perceived overall risk was modeled as a second-order construct having the four significant risk facets mentioned in the Research model and hypotheses development section above as first-order components. Integrated value-risk perception resulted as a third-order latent variable and this assumption is concordant with similar research on the role of perceived value in ICT adoption (Turel et al. 2010). Latent variables were measured through several items adapted from questionnaires validated by previous research in information systems (Featherman and Pavlou 2003, Kim et al. 2007, Turel et al. 2010, Venkatesh and Davis 2000, Venkatesh et al. 2002) and consumer behavior (Laroche et al. 2003, Stone and Grønhaug 1993). Responses were collected on 7-point Likert scales. Measurement items for the model constructs are presented in Appendix A. After eliminating the responses with uncommon pattern of the answers (e.g., same answer number for all item questions), a number of 289 valid cases out of the 300 complete answers were recorded. The 289 valid cases (148 coming from the ‘Pros’ subsample and 141 from the ‘Cons’ subsample, respectively) were retained for subsequent analyses. Respondents were 50.17% female and 49.83% male, reporting an average age of 45.01years. Their average experience with smartphones was 11.03 years. Participants reported using their smartphones 36.00 minutes per day for speaking or texting and 23.23 minutes per day for 18

browsing the Internet, on average. The largest part of their daily activities on smartphones resulted to be for entertainment followed at a significant distance by banking, shopping and business at about the same level. Respondents were using 2.28 credit cards at the time of the survey and were having 21.12 years experience with credit cards, on average. The sum of their credit card bills for the previous month amounted to $1,679.03, on average.

5.1 Measurement model A preliminary step in data analysis was to test on the potential influence of common method variance (CMV) as all variables (both exogenous and endogenous) were collected in the same survey (Sharma et al. 2009). For that, a Harman’s one-factor test was conducted according to guidelines from Podsakoff et al. (2003). Thus, all items pertaining to the theoretical model were entered into an exploratory factor analysis. The unrotated solution produced by SPSS resulted in four factors with eigenvalues greater than one, the smallest of these being 1.538. The first factor accounted for only 33.0% of the variance while all factors isolated through this method explained 75.5% of the variance. These results indicate that the variables in the model do not load on a single factor. Furthermore, a second test for detecting potential CMV was a visual inspection of the correlation matrix of first-order factors. As Table 3 below shows, all absolute values of factor correlations were below 0.9 hence CMV was not considered a concern (Pavlou et al. 2007). Following this analysis, the measurement model was assessed with PLS-SEM by running SmartPLS (Ringle et al. 2005). After the first run of the program it became necessary to drop one item (pertaining to the Perceived Social Risk factor) out of the 25 of the measurement model because it displayed poor loading. SmartPLS was run again and this time the output revealed appropriate measurement outcomes. Thus, as Table 1 shows, all first-order constructs had Average Variance Extracted (AVE) values above 0.5 as well as Composite Reliability and Cronbach’s 19

alpha values above 0.7. As indicated by Table 2, all factor loadings were above the 0.7 threshold, all item standard deviation and standard error values were relatively small and all items were significant at a level better than 0.001. Therefore, results in Tables 1 and 2 led to the conclusion that the measurement model was appropriate in terms of reliability and convergent validity (Fornell and Larcker 1981, Jarvenpaa et al. 2004).

Table 1 Average Variance Extracted (AVE), Composite reliability, and Cronbach’s alpha values for firstorder constructs Construct Utilitarian value Enjoyment value Social value Time risk Social risk Psychological risk Privacy risk Behavioral intention

AVE 0.696 0.864 0.918 0.836 0.806 0.887 0.872 0.963

Composite Reliability 0.901 0.950 0.978 0.939 0.892 0.959 0.953 0.981

Cronbach’s Alpha 0.852 0.921 0.970 0.901 0.759 0.936 0.926 0.962

20

Table 1 Factor loading and t-statistic levels for first-order constructs Mean Item and factor USE1 <- Utilitarian value USE2 <- Utilitarian value USE3 <- Utilitarian value USE4 <- Utilitarian value ENJ1 <- Enjoyment value ENJ2 <- Enjoyment value ENJ3 <- Enjoyment value SOC1 <- Social value SOC2 <- Social value SOC3 <- Social value SOC4 <- Social value PTR1 <- Time risk PTR2 <- Time risk PTR3 <- Time risk PSR1 <- Social risk PSR2 <- Social risk PPSYR1 <- Psychological risk PPSYR2 <- Psychological risk PPSYR3 <- Psychological risk PPR1 <- Privacy risk PPR2 <- Privacy risk PPR3 <- Privacy risk BI1 <- Behavioral intention BI2 <- Behavioral intention

4.20 4.54 3.58 4.11 3.84 4.02 3.58 2.90 2.79 2.76 2.68 3.25 3.44 3.39 3.49 2.89

Standard deviation 1.55 1.57 1.54 1.52 1.53 1.54 1.64 1.68 1.70 1.67 1.63 1.70 1.58 1.64 1.84 1.75

Factor loading 0.754 0.863 0.789 0.921 0.941 0.937 0.911 0.933 0.970 0.962 0.966 0.896 0.941 0.904 0.907 0.888

Standard error 0.191 0.212 0.187 0.223 0.189 0.187 0.182 0.157 0.169 0.173 0.167 0.035 0.015 0.024 0.025 0.041

t-Statistic 3.946 4.077 4.211 4.123 4.983 5.013 5.000 5.939 5.745 5.552 5.774 25.351 61.126 37.556 35.840 21.770

4.37

1.84

0.909

0.031

29.359

3.91

1.80

0.962

0.012

81.543

3.84 4.39 4.51 4.78

1.76 1.71 1.75 1.77

0.955 0.930 0.954 0.916

0.012 0.018 0.012 0.025

77.603 52.494 82.551 36.160

3.95

1.65

0.982

0.006

159.557

3.95

1.67

0.981

0.007

150.817

The next test of the measurement model consisted in examining visually a matrix having the square root of the AVEs for first-order factors on the diagonal and the correlation with the other first-order factors, as calculated by SmartPLS, off diagonal (Table 3). As diagonal numbers were larger than all the other numbers on their corresponding rows and columns, we drew the conclusion that the measurement model manifested appropriate discriminant validity too, 21

according to literature recommendations (Gefen and Straub 2005). This conclusion is strengthened by another test relying on the table of loadings and cross-loadings for all first-order constructs produced by SmartPLS (Table 4). As in almost all cases item loadings were larger on their corresponding constructs than on the other constructs, the model has appropriate discriminant validity (Bontis 2004, Gefen and Straub 2005). Since reliability together with convergent and discriminant (i.e., construct) validity tests met or exceeded the thresholds specified in relevant literature, the measurement model was considered appropriate, thus allowing the next step of the PLS analysis – evaluation of the structural model.

Table 3 Square-root AVEs (on diagonal) and correlation coefficients (off diagonal) for first-order constructs.

Enjoyment value Enjoyment value Behavioral intention Privacy risk Psychological risk Social risk Social value Time risk Utilitarian value

Behavioral intention

Privacy risk

Psychological risk

Social risk

Social value

Time risk

0.770 -0.167

0.981 -0.286

0.934

-0.255 0.060 0.424 -0.156

-0.360 -0.029 0.383 -0.238

0.742 0.315 -0.144 0.412

0.942 0.373 0.031 0.524

0.898 0.307 0.557

0.958 0.218

0.914

0.807

0.746

-0.199

-0.283

0.033

0.334

-0.207

Utilitarian value

0.929

0.834

Table 4 Item loadings and cross-loadings for first-order constructs. Enjoyment value (ENJ)

Behavioral intention (BI)

Privacy risk (PPR)

Psychological Social risk risk (PPSYR) (PSR)

UtilitaSocial Time rian value risk value (SOC) (PTR) (USE) 22

BI1 BI2 ENJ1 ENJ2 ENJ3 PPR1 PPR2 PPR3 PPSYR1 PPSYR2 PPSYR3 PSR1 PSR2 PTR1 PTR2 PTR3 SOC1 SOC2 SOC3 SOC4 USE1 USE2 USE3 USE4

0.761 0.749 0.941 0.937 0.911 -0.165 -0.192 -0.105 -0.307 -0.193 -0.220 0.039 0.070 -0.126 -0.124 -0.177 0.438 0.407 0.390 0.387 0.579 0.695 0.619 0.783

0.982 0.981 0.729 0.708 0.709 -0.272 -0.286 -0.242 -0.403 -0.295 -0.319 -0.036 -0.015 -0.241 -0.179 -0.234 0.432 0.356 0.339 0.336 0.511 0.605 0.593 0.754

-0.293 -0.268 -0.161 -0.150 -0.152 0.930 0.954 0.916 0.790 0.675 0.631 0.349 0.211 0.347 0.392 0.391 -0.120 -0.129 -0.158 -0.146 -0.081 -0.102 -0.246 -0.214

-0.360 -0.346 -0.239 -0.257 -0.215 0.712 0.696 0.669 0.909 0.962 0.955 0.356 0.311 0.471 0.470 0.495 0.019 0.048 0.020 0.035 -0.142 -0.271 -0.183 -0.324

-0.030 -0.027 0.036 0.056 0.075 0.352 0.281 0.245 0.339 0.359 0.356 0.907 0.888 0.578 0.490 0.462 0.261 0.309 0.298 0.309 0.084 -0.025 0.109 -0.041

0.386 0.366 0.389 0.349 0.446 -0.067 -0.165 -0.177 -0.079 0.091 0.077 0.173 0.386 0.242 0.195 0.161 0.933 0.970 0.962 0.966 0.236 0.133 0.476 0.265

-0.224 -0.243 -0.128 -0.175 -0.131 0.465 0.399 0.280 0.428 0.515 0.538 0.450 0.555 0.896 0.941 0.904 0.172 0.229 0.216 0.219 -0.119 -0.237 -0.061 -0.257

0.730 0.734 0.803 0.767 0.678 -0.199 -0.209 -0.145 -0.304 -0.244 -0.250 0.046 0.012 -0.201 -0.160 -0.207 0.375 0.313 0.302 0.284 0.754 0.863 0.789 0.921

5.2 Structural model We ran SmartPLS with bootstrap with 200 re-samples to obtain path coefficients, their significance levels as well as the coefficients of determination. Relevant results are presented in Table 5 and Figure 2.

Table 5 Hypotheses testing results. Hypothesis H1 H2

Theoretical model path Utilitarian value -> Integrated value-risk Enjoyment value ->

Coefficient

Standard error

t-Statistic

p-Value

0.355 0.337

0.043 0.056

8.2629 6.0041

<0.001 <0.001

Test outcome Supported 23

H3 H4a H4b H4c H4d H5 H6

Integrated value-risk Social value -> Integrated value-risk Time risk -> Overall risk Social risk -> Overall risk Psychological risk -> Overall risk Privacy risk -> Overall risk Overall risk -> Integrated value-risk Integrated value-risk -> Behavioral intention

Supported 0.279 0.302 0.127

0.095 0.036 0.033

2.9348 8.376 3.7936

<0.01 <0.001 <0.001

Supported Supported Supported Supported

0.411 0.386

0.032 0.039

12.8134 9.8729

<0.001 <0.001

Supported

-0.459

0.199

2.3035

<0.05

Supported

0.786

0.186

4.2116

<0.001

Supported

Fig. 2. Structural evaluation of the theoretical adoption model. Significance levels: * = 0.05; ** = 0.01; *** = 0.001 24

A visual inspection of Table 5 and Figure 2 shows all hypotheses proposed were supported. Significance levels were 0.001 or better with two exceptions: the influence of Social Value in the Perceived Value-Risk (significant at 0.01) and the influence of Overall Risk in the Perceived Value-Risk (significant at 0.05). The model had moderately high explanatory power capturing 61.7% of the variability of the intention to use NFC smartphones for credit card payments. As all of the paths in the model proved to be statistically significant and as the coefficient of determination of the endogenous construct was moderately high for the information systems domain, the model was considered to be reasonably good, according to relevant literature (Bontis et al. 2002). SmartPLS also calculated the total (direct and indirect) effects of the first-order factors on the behavioral intention, as shown in Table 6. Total effect coefficients confirm the positive influence of the ‘gain’ side of the value perception (utilitarian and enjoyment, especially) and negative influence of the ‘cost’ side of the value (psychological and privacy risks, in particular).

Table 6 Total effects of first-order factors on the behavioral intention. Factor Utilitarian value Enjoyment value Social value Time risk Social risk Psychological risk Privacy risk

Total effect coefficient 0.279 0.265 0.219 -0.109 -0.046 -0.148 -0.139

t-Statistic 3.943 4.384 3.156 2.805 2.503 3.128 3.201

p-Value <0.001 <0.001 0.002 0.005 0.013 0.002 0.002

5.3 Control variables All demographic factors measured were tested as potential control variables. These were included in the theoretical model in turn with paths to the Perceived Value-Risk and to the 25

Behavioral Intention. Possible changes in the measurement model and significance values for the new added paths together with the variance explained (R2) value for the endogenous construct were recorded each time (Table 7).

Table 7 Control variables testing results.

Control variable Uncontrolled model ‘Pros’ or ‘Cons’ subsample Gender Age Smartphone experience Time of daily speaking or texting Time of daily browsing the Internet Number of credit cards used Credit card experience Sum of latest credit card bills

t-Statistic of path to Behavioral intention 0.685 0.493 0.694 0.751 0.327 0.178 0.730 0.193 0.568

Variance explained for Behavioral intention 0.617 0.619 0.618 0.619 0.619 0.617 0.617 0.619 0.617 0.619

The potential control variable of main interest was the subsample the respondents fell in depending on the information they were offered before completing the survey – i.e., ‘Pros’ or ‘Cons’ the new ICT. This variable did not appear to affect the model – its path to Behavioral Intention was not significant at the 0.05 statistical level or better and the variable caused a minor change of the coefficient of determination from 0.617 to 0.619 without any alteration of the measurement model. The same variable capturing subsample type was also tested for a possible moderation effect of the Integrated Value-Risk over the Behavioral Intention. This effect was not significant either at the 0.05 level. Control variable tests for gender and age, smartphone experience and pattern of use (i.e., time of speaking or texting, or browsing the Internet daily) and credit card experience and features of usage (i.e., number of credit cards used and sum of latest credit card bills) did not reveal any significant path to the endogenous construct, as Table 7 shows. A minor increase of the coefficient 26

of determination, with less than 0.5%, was recorded in some cases and no changes in the measurement model were noticed.

6. Discussion and conclusions This study is an empirical investigation of the adoption of contactless mobile payments with NFC enabled smartphones through a perceived value-risk approach. To investigate this issue the study proposed the research question: What is the combined effect of perceived user value and risk on the adoption intention of contactless smartphone payments? In order to answer this question the study constructed an unbiased perceived value-risk model contrasting the gain and the cost of using the ICT service. As the major theoretical contribution, this study conceptualizes an innovative third-order factor, called integrated value-risk perception, that captures in an unbiased approach the benefits and sacrifices of adopting NFC mobile payments from a consumer perspective. Thus, while the ‘gain’ side of the value-risk is captured through utilitarian, enjoyment and social benefits, the ‘cost’ side is expressed through a secondorder perceived risk factor that captures consumer possible doubts in several directions. With this approach the study is significantly different from previous research in the following key points: 1. It uses a perceived value framework to investigate a mobile banking (i.e., service) adoption. Thus, recent comprehensive literature reviews of mobile banking (Shaikh and Karjaluoto 2015) and of mobile data services (Ovčjak et al. 2015) adoption did not identify perceived value (or benefit-cost) approaches among the frequently used theoretical designs. Therefore, this study answers calls of previous research asking to investigate consumer ebanking adoption in general with less frequent but potentially useful theoretical and methodological approaches (Hoehle et al. 2012). Moreover, this research also answers the 27

call to use novel approaches to investigate the meaning and role of consumer value in behavioral research (Gallarza et al. 2011). 2. Compared to the relatively infrequent research which used a perceived value framework to investigate ICT adoption, this study takes into account cost factors in addition to the popular benefit factors. Thus, most of the value-based approaches in the available literature considered only the one-sided perspective of the benefit (or gain) aspects of value that determine adoption– e.g., Al-Debei and Al-Lozi (2014), Chen et al. (2008), Hsiao et al. (2016), Kim and Han (2009), and Yang and Lin (2014). 3. As a refinement compared to the relatively few studies that took into account the cost (or sacrifice) side of value in the ICT adoption equation, this research integrates a multidimensional perceived risk (i.e., fears of wasting time, of being disapproved by significant other people, of having overall doubts, and of facing privacy threats if using contactless payments with NFC smartphones) into a multi-sided perceived value. This is considered a more granular analysis than looking at predominantly monetary aspects (e.g., Kim et al. (2007), Lin et al. (2012), Turel et al. (2007), and Turel et al. (2010)) or at a lumped risk perception external to value (e.g., Chiu et al. (2014), and Xu et al. (2015)). The model was tested in an empirical research through a cross-sectional survey conducted with 289 respondents in Canada. Results showed that all ‘gain’ facets were significant for the perceived value-risk, with the utilitarian and hedonic sides having the largest weights (0.355 and 0.337, respectively, both significant at a level below 0.001, as Table 5 and Figure 2 show). This is consistent with previous research investigating ICT adoption through a perceived value lens (Rintamäki et al. 2006, Turel et al. 2007, Turel et al. 2010) or a motivational lens (Davis et al. 1992). Social side of the value, although still significant, has a comparatively lower importance that could be explained by the fact that credit card use is usually done in an individual context. 28

Another theoretical contribution of this research is the examination at a more detailed level of the cost side of the value, that has been comparatively less investigated in previous value-based models in IS research (Xu et al. 2015). Thus the sacrifice side of the value was captured here through a multifaceted perceived risk that proved to have a negative effect on the overall value seen by consumers and, consequently, on the intention to use the ICT application, as Table 5 and Figure 2 show. This is consistent with previous research in consumer behavior that demonstrated that perceived risk influences negatively perceived value (Beneke et al. 2013) as well as with research in IS that showed that risk perception is a deterrent to ICT adoption (Cocosila et al. 2009, Cocosila and Archer 2010, Featherman and Pavlou 2003). As Figure 2 and Table 6 indicate, all four risk facets considered in the theoretical model were significant at a statistical level of 0.05 or better but all had a comparatively lower total effect on the intention to use the ICT than the ‘gain’ factors. Perceived psychological risk and perceived privacy risk proved to be the most important ‘cost’ factors (with a total effect of -0.148 and -0.139, respectively, on the intention to use, as Table 6 shows). Thus, this research also answered calls from previous literature to better scrutinize risk effects in value-based models (Snoj et al. 2004). Overall, the innovative value-risk integrated perception introduced by this study can be considered as valid since all the paths leading directly or indirectly to the third-order construct are significant at a statistical level of 0.05 or better and have reasonably large weight values, as Table 5 and Figure 2 show. Furthermore, the integrated value-risk construct explained 61.7% of the consumers’ intent to use NFC mobile services and this reinforces the validity of both the new third-order construct and of the whole theoretical model developed by this study. As the major applied contribution of this study, findings depicted in Figure 2 and Tables 5 and 6 may give practitioners an understanding of customers’ value and risk priorities which can be used to design appropriate success strategies. Thus, the results indicate that consumers see more 29

benefits than threats in the use of NFC smartphones for payments because the gain factors have higher significance levels and larger total effect values than the sacrifice factors. These findings show that developers and promoters of credit card payments through NFC smartphones should first reinforce consumer perceptions of value (utility and enjoyment, especially) in order to increase adoption and use. However, as sacrifice factors play a non-negligible role in the overall value-risk combination, concerns on the justification for the service and possible privacy issues have to be mitigated in order to increase the success rate of this service. We found that none of the demographic characteristics influenced consumer perceptions on the value-risk of the ICT service and their intention to use, as summarized in Table 7. Remarkably, the initial contrasting information provided to the two subsamples of respondents (i.e., NFC smartphone payments have many advantages or may pose problems, respectively) did not have a significant effect on the results. The considerable experience of the respondents with both smartphones (11.03 years, on average) and credit cards (21.12 years, on average) may be an explanation for the lack of difference between the views of the ‘Pros’ and ‘Cons’ subsamples. Therefore, the reinforcing and mitigation measures suggested above should not differentiate among various categories of consumers. Like virtually any empirical research in information systems, this study had also some limitations. Participants were recruited from the tens of thousands respondents pre-registered Canada-wide with a market survey company and meeting the including conditions. Participants self-selected and were admitted to the study until the quota of 300 complete answers was met. Although sample size and representativeness may often be debatable in an empirical research, the sample of this study was larger than the minimal value recommended by PLS-SEM literature (Bontis 1998, Chin 1998b, Jarvenpaa et al. 2004) and was considered to be realistic as it was drawn from the general population of a large country. However, the model proposed by this study 30

should be confirmed in future research with other samples since perceived value may have different effects on consumer behaviors in different contexts (Hsiao and Chen 2016). Further research should also investigate why none of the demographics, the contradictory information on a new and sensitive technology in particular, had an influence on the theoretical model. A more granular analysis considering the type of smartphone activity and credit card payment the respondents use more frequently may be necessary. Furthermore, a contrast between consumers who already used NFC smartphone payments and those who did not use yet may be considered for future research. In general, it could be concluded that the limitations did not affect the validity of this study but, rather, offered ideas for future research. Overall, this research is a first attempt to investigate in a comprehensive approach consumer perceptions on the potential use of credit card payments with Near Field Communication smartphones. In order to reach this goal, as a major contribution to electronic commerce research, this study developed a theoretical model relying on an innovative factor integrating a multidimensional perceived risk construct into a multi-sided perceived value factor. Thus, this research proposed a broader view of the duality Gain-Cost in the perception of value of an ICT artifact that may be used in similar adoption studies.

Acknowledgement The authors are grateful to the Editor-in-Chief, Associate Editor and three anonymous reviewers of this journal for their very valuable feedback and recommendations. This study is part of a larger project supported through a research award by Athabasca University.

References

31

ABI Research Technology Market Intelligence., 2015. NFC Mobile Payments. URL: http://www.businesswire.com/news/home/20121018006048/en/NFC-Mobile-PaymentTransaction-Spend-Hit-100 (Accessed on November 4, 2015). Al-Debei, M. M., Al-Lozi, E., 2014. Explaining and predicting the adoption intention of mobile data services: A value-based approach. Computers in Human Behavior, 35, 326-338. Agarwal, S., Teas, R. K., 2001. Perceived value: mediating role of perceived risk. Journal of Marketing theory and Practice, 9, 4, 1-14. Agarwal, S., Teas, R. K., 2004. Cross-national applicability of a perceived risk-value model. Journal of Product & Brand Management, 13, 4, 242-256. Bagadia, P., & Bansal, A. (2016). Risk perception and adoption of mobile banking services: A review. IUP Journal of Information Technology, 12, 1, 52-71. Beneke, J., Flynn, R., Greig, T., Mukaiwa, M., 2013. The influence of perceived product quality, relative price and risk on customer value and willingness to buy: A study of private label merchandise. Journal of Product & Brand Management, 22, 3, 218-228. Bontis, N., 1998. Intellectual capital: An exploratory study that develops measures and models. Management Decision, 36, 2, 63-76. Bontis, N., 2004. National intellectual capital index: A United Nations initiative for the Arab region. Journal of Intellectual Capital, 5, 1, 13-39. Bontis, N., Crossan, M. M., Hulland, J., 2002. Managing an organizational learning system by aligning stocks and flows. Journal of Management Studies, 39, 4, 437-469. Burgelman, R., Christensen, C., Wheelwright, S., 2004. Strategic Management of Technology and Innovation. Boston, USA: McGraw-Hill Irwin.

32

Canada Newswire., 2015. CIBC and Rogers complete the first mobile credit card transaction in Canada. URL: http://www.newswire.ca/news-releases/cibc-and-rogers-complete-the-firstmobile-credit-card-transaction-in-canada-511093801.html (Accessed on November 10, 2015). Chang, E. C., Tseng, Y. F., 2013. Research note: E-store image, perceived value and perceived risk. Journal of Business Research, 66, 7, 864-870. Chen, Y. C., Shang, R. A., Lin, A. K., 2008. The intention to download music files in a P2P environment: Consumption value, fashion, and ethical decision perspectives. Electronic Commerce Research and Applications, 7, 4, 411-422. Chin, W. W., 1998a. Commentary: Issues and opinion on structural equation modeling. MIS Quarterly, 22, 1, vii-xvi. Chin, W. W., 1998b. The partial least squares approach for structural equation modeling, in Marcoulides, G. A. (Ed), Modern methods for business research. Methodology for business and management, (pp. 295-336), Mahwah, NJ, USA: Lawrence Erlbaum Associates Publishers. Chiu, C. M., Wang, E. T., Fang, Y. H., Huang, H. Y., 2014. Understanding customers' repeat purchase intentions in B2C e-commerce: The roles of utilitarian value, hedonic value and perceived risk. Information Systems Journal, 24, 1, 85-114. Chwelos, P., Benbasat, I., Dexter, A. S., 2001. Research report: Empirical test of an EDI adoption model. Information Systems Research, 12, 3, 304-321. Cocosila, M., Archer, N., Yuan, Y., 2009. Early investigation of new information technology acceptance: A perceived risk - motivation model. Communications of the Association for Information Systems, 25, 30.

33

Cocosila, M., Archer, N., 2010. Adoption of mobile ICT for health promotion: An empirical investigation. Electronic Markets - The International Journal on Networked Business, 20, 3-4, 241-250, doi: 10.1007/s12525-010-0042-y. Dahlberg, T., Guo, J., Ondrus, J., 2015. A critical review of mobile payment research, Electronic Commerce Research and Applications, 14, 5, 265-284. Davis, F., 1989. Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Quarterly, 13, 3, 319-340. Davis, F., Bagozzi, R., Warshaw, P., 1992. Extrinsic and intrinsic motivation to use computers in the workplace. Journal of Applied Social Psychology, 22, 14, 1111-1132. De Reuver, M., Verschuur, E., Nikayin, F., Cerpa, N., Bouwman, H., 2015. Collective action for mobile payment platforms: A case study on collaboration issues between banks and telecom operators. Electronic Commerce Research and Applications, 14, 5, 331-344. Featherman, M. S., Pavlou, P., 2003. Predicting e-services adoption: A perceived risk facets perspective. International Journal of Human Computer Studies, 59, 1, 451-474. Fishbein, M., Ajzen, I., 1975. Belief, attitude, intention and behavior: An introduction to theory and research. Reading, Massachusetts: Addison-Wesley Publishing Company. Fornell, C., Larcker, D. F., 1981. Evaluating structural equation models with unobservable variables and measurement error. Journal of Marketing Research, 18, 1, 39-50. Forsythe, S. M., Shi, B., 2003. Consumer patronage and risk perceptions in Internet shopping. Journal of Business Research, 56, 11, 867-875. Gallarza, M. G., Gil‐Saura, I., Holbrook, M. B., 2011. The value of value: Further excursions on the meaning and role of customer value. Journal of Consumer Behaviour, 10, 4, 179-191. Gefen, D., Straub, D., 2005. A Practical guide to factorial validity using PLS-graph: Tutorial and annotated example. Communications of the Association for Information Systems, 16, 91-109. 34

Gefen, D., Straub, D. W., Rigdon, E. E., 2011. An update and extension to SEM guidelines for administrative and social science research. MIS Quarterly, 35, 2, iii-xiv. Goodhue, D. L., Lewis, W., Thompson, R., 2012a. Does PLS have advantages for small sample size or non-normal data? MIS Quarterly, 36, 3, 891-1001. Goodhue, D. L., Lewis, W., Thompson, R., 2012b. Comparing PLS to regression and LISREL: A response to Marcoulides, Chin, and Saunders. MIS Quarterly, 36, 3, 703-716. Grassie, K., 2007. Easy handling and security make NFC a success. Card Today Technology, 19, 10, 12-13. Hoehle, H., Scornavacca, E., Huff, S., 2012. Three decades of research on consumer adoption and utilization of electronic banking channels: A literature analysis. Decision Support Systems, 54, 1, 122-132. Hsiao, K. L., Chen, C. C., 2016. What drives in-app purchase intention for mobile games? An examination of perceived values and loyalty. Electronic Commerce Research and Applications, 16, 18-29. Hsiao, C. H., Chang, J. J., Tang, K. Y., 2016. Exploring the influential factors in continuance usage of mobile social Apps: Satisfaction, habit, and customer value perspectives. Telematics and Informatics, 33, 2, 342-355. Hsu, C. L., Lin, J. C. C., 2015. What drives purchase intention for paid mobile apps?–An expectation confirmation model with perceived value. Electronic Commerce Research and Applications, 14, 1, 46-57. Jarvenpaa, S. L., Shaw, T. R., Staples, D. S., 2004. Toward contextualized theories of trust: The role of trust in global virtual teams. Information Systems Research, 15, 3, 250-267. Jung, Y., 2014. What a smartphone is to me: Understanding user values in using smartphones. Information Systems Journal, 24, 4, 299-321. 35

Kim, B., Han, I. (2009). What drives the adoption of mobile data services? An approach from a value perspective. Journal of Information Technology, 24, 1, 35-45. Kim, D. J., Ferrin, D. L., Rao, H. R., 2008. A trust-based consumer decision-making model in electronic commerce: The role of trust, perceived risk, and their antecedents. Decision Support Systems, 44, 2, 544-564. Kim, H. W., Chan, H. C., Gupta, S., 2007. Value-based adoption of mobile Internet: An empirical investigation. Decision Support Systems, 43, 111-126. Kim, H. W., Gupta, S., Koh, J., 2011. Investigating the intention to purchase digital items in social networking communities: A customer value perspective. Information & Management, 48, 6, 228-234. Kim, Y., Han, H. J., 2008. The effects of perceived risk and technology type on users’ acceptance of technologies. Information & Management, 45, 1-9. Kuo, Y. F., Wu, C. M., Deng, W. J., 2009. The relationships among service quality, perceived value, customer satisfaction, and post-purchase intention in mobile value-added services. Computers in Human Behavior, 25, 4, 887-896. Lapointe, L., Rivard, S., 2005. A multilevel model of resistance to information technology implementation. MIS Quarterly, 29, 3, 461-491. Laroche, M., Bergeron, J., Goutaland, C., 2003. How intangibility affects perceived risk: The moderating role of knowledge and involvement. Journal of Services Marketing, 17, 2, 122-140. Lee, M. R., Yen, D. C., Hsiao, C. Y., 2014. Understanding the perceived community value of Facebook users. Computers in Human Behavior, 35, 350-358.

36

Lee, W.-H., Hwang, K.-P., Lin, W.-J., Shieh, J.-L., 2015. An analysis of trusted service manager development modes by mobile operating system designers in Taiwan. Electronic Commerce Research and Applications, 14, 6, 592-602. Li, J., Liu, J.-L., Ji, H.-Y., 2014. Empirical study of influence factors of adaption intention of mobile payment based on TAM model in China. International Journal of u- and e- Service, Science and Technology, 7, 1, 119-132. Lim, N., 2003. Consumers’ perceived risk: Sources versus consequences. Electronic Commerce Research and Applications, 2, 216-228. Lin, C. H., Sher, P. J., Shih, H. Y., 2005. Past progress and future directions in conceptualizing customer perceived value. International Journal of Service Industry Management, 16, 4, 318336. Lin, T. C., Wu, S., Hsu, J. S. C., Chou, Y. C., 2012. The integration of value-based adoption and expectation–confirmation models: An example of IPTV continuance intention. Decision Support Systems, 54, 1, 63-75. Lohmoller, J. B., 1989. Latent Variable Path Modelling with Partial Least Squares. Heidelberg, Germany: Physica-Verlag. NFC World Forum., 2015. NFC phones: The definitive list. URL: http://www.nfcworld.com/nfcphones-list (Accessed on November 20, 2015). Ou, C. X., Pavlou P., Davison, R., 2014.Swift Guanxi in online marketplaces: The role of computer-mediated communication technologies. MIS Quarterly, 38, 1, 209-A24. Ovčjak, B., Heričko, M., Polančič, G., 2015. Factors impacting the acceptance of mobile data services–A systematic literature review. Computers in Human Behavior, 53, 24-47. Park, B. W., Lee, K. C., 2011. Exploring the value of purchasing online game items. Computers in Human Behavior, 27, 6, 2178-2185. 37

Pavlou, P. A., 2003. Consumer acceptance of electronic commerce - Integrating trust and risk with the Technology Acceptance Model. International Journal of Electronic Commerce, 7, 3, 101134. Pavlou, P. A., Liang, H. G., Xue, Y. J., 2007. Understanding and mitigating uncertainty in online exchange relationships: A principal-agent perspective. MIS Quarterly, 31, 1, 105-136. Petter, S., Straub, D., Rai, A., 2007. Specifying formative constructs in information systems research. MIS Quarterly, 31, 4, 623-656. Podsakoff, P. M., MacKenzie, S. B., Lee, J.-Y., Podsakoff, N. P., 2003. Common method biases in behavioral research: A critical review of the literature and recommended remedies. Journal of Applied Psychology, 88, 5, 879-903. Rezaei, S., Ghodsi, S. S., 2014. Does value matters in playing online game? An empirical study among massively multiplayer online role-playing games (MMORPGs). Computers in Human Behavior, 35, 252-266. Ringle C. M., Wende, M. S., Will, S., 2005. SmartPLS 2.0 (M3) Beta. URL: http://www.smartpls.de (Accessed on October 08, 2015). Ringle, C. M., Sarstedt, M., Straub, D., 2012. A critical look at the use of PLS-SEM in MIS Quarterly. MIS Quarterly, 36, 1, iii-xiv. Rintamäki T., Kanto, A., Spence, M. T., 2006. Decomposing the value of department store shopping into utilitarian, hedonic, and social dimensions. International Journal of Retail and Distribution Management, 34, 1, 6-24. Rose, C., 2012. Close contact: An examination of the future of near field communications. International Journal of Management & Information Systems, 16, 1, 95-100. Shaikh, A. A., Karjaluoto, H., 2015. Mobile banking adoption: A literature review. Telematics and Informatics, 32, 1, 129-142. 38

Sharma, R., Yetton, P., Crawford, J., 2009. Estimating the effect of common method variance: The method-method pair technique with an illustration from TAM research. MIS Quarterly, 33, 3, 473-490. Sheth J. D., Newman, B. I., Gross, B. I., 1991. Consumption Values and Market Choice. Cincinnati, USA: South Western Publishing. Snoj, B., Korda, A. P., Damijan, M., 2004. The relationships among perceived quality, perceived risk and perceived product value. Journal of Product & Brand Management, 13, 2/3, 156-167. Soper, D., 2015. A-priori sample size calculator for structural equation models. URL: http://www.danielsoper.com/statcalc3/calc.aspx?id=89 (Accessed on December 28, 2015) Stone, R. N., Grønhaug, K., 1993. Perceived risk: Further considerations for the marketing discipline. European Journal of Marketing, 27, 3, 39-50. Sweeney, J. C., Soutar, G. N., 2001. Consumer perceived value: The development of a multiple item scale. Journal of Retailing, 77, 2, 203-220. Sweeney, J. C., Soutar, G. N., Johnson, L.W., 1999. The role of perceived risk in the quality-value relationship: A study in a retail environment. Journal of Retailing, 75, 1, 77-105. Thomas, R. D., Lu, I. R. R., Cedzynski, M., 2005. Partial Least Squares: A critical review and a potential alternative. Administrative Sciences Association of Canada (ASAC) Conference, Toronto, Ontario, Canada. Turel, O., Serenko, A., Bontis, N., 2007. User acceptance of wireless short messaging services: Deconstructing perceived value. Information & Management, 44, 63-73. Turel, O., Serenko, A., Bontis, N., 2010. User acceptance of hedonic digital artifacts: A theory of consumption values perspective. Information & Management, 47, 53-59.

39

Venkatesh, V., Davis, F., 2000. A theoretical extension of the Technology Acceptance Model: Four longitudinal field studies. Management Science, 46, 2, 186-204. Venkatesh, V., Morris, M. G., Davis, G. B., Davis, F. D., 2003. User acceptance of information technology: Toward a unified view. MIS Quarterly, 27, 3, 425-478. Venkatesh, V., C. Speier, Morris, M. G., 2002. User acceptance enablers in individual decision making about technology: Toward an integrated model. Decision Sciences, 33, 2, 297-316. Visa Pay Wave., 2015. URL: http://www.visa.ca/merchant/products-and-services/paywave/index.jsp (Accessed November 20, 2015). Xu, C., Peak, D., Prybutok, V., 2015. A customer value, satisfaction, and loyalty perspective of mobile application recommendations. Decision Support Systems, 79, 171-183. Yang, H. L., Lin, C. L., 2014. Why do people stick to Facebook web site? A value theory-based view. Information Technology & People, 27, 1, 21-37. Yang, Q., Pang, C., Liu, L., Yen, D. C., Tarn, J. M., 2015. Exploring consumer perceived risk and trust for online payments: An empirical study in China’s younger generation. Computers in Human Behavior, 50, 9-24. Yang, H., Yu, J., Zo, H., Choi, M., 2016. User acceptance of wearable devices: An extended perspective of perceived value. Telematics and Informatics, 33, 2, 256-269. Yu, A. W., Prybutok, V. R., Koh, C. E., Hanus, B., 2012. A nomological model of RFID privacy concern. Business Process Management Journal, 18, 3, 420-444. Zeithaml, V. A., 1988. Consumer perceptions of price, quality and value: A means-end model and synthesis of evidence. Journal of Marketing, 52, 2-22.

40

Zhu, Y., Chen H., 2011. An integrative model of consumer’s adoption of RFID credit card service. Proceedings of the 4th Hawaii International Conference on System Sciences, 1-9, doi: 10.1109/HICSS.2011.57.

Appendix A Theoretical model first-order constructs and corresponding measurement items

Perceived Social Risk •

My friends and colleagues’ negative opinions about my signing up for the contactless smartphone payments service would cause me to feel concern.

•

If signing up for the contactless smartphone payments service, I would be concerned about what people whose opinion is of value for me would think of me, if I made a bad choice.

•

My subscribing to the contactless smartphone payments service would cause me concern about what my friends would think of me, if I made a bad choice.

Perceived Time Risk •

Using the contactless smartphone payments service could lead to an inefficient use of my time.

•

Using the contactless smartphone payments service could require more time than when not using them.

•

The demands on my schedule are such that using the contactless smartphone payments service concerns me because it could create even more time pressures on me that I don’t need.

Perceived Privacy Risk

41

•

My use of the contactless smartphone payments service would cause me to lose control over the privacy and confidentiality of my personal information.

•

Signing up for and using the contactless smartphone payments service would lead to a loss of privacy for me because my personal information could be used without my knowledge.

•

Internet hackers (criminals) might take control of my information if I used the contactless smartphone payments service.

Perceived Psychological Risk •

The thought of signing up for the contactless smartphone payments service makes me feel uncomfortable.

•

The thought of signing up for the contactless smartphone payments service gives me an unwanted feeling of anxiety.

•

The thought of signing up for the contactless smartphone payments service causes me to experience unnecessary tension.

Utilitarian Value •

Using the contactless smartphone payments service would help me save time and efforts.

•

Using the contactless smartphone payments service would be convenient.

•

Using the contactless smartphone payments service would help me to better manage my expenses.

•

I found the contactless smartphone payments service would be useful overall.

Enjoyment Value •

I found the contactless smartphone payments service would be enjoyable.

•

The actual process of using the contactless smartphone payments service would be pleasant.

42

•

I would have fun using the contactless smartphone payments service.

Social Value •

The use of the contactless smartphone payments service would help me feel acceptable among my friends.

•

The use of the contactless smartphone payments service would improve the way I am perceived by my peers.

•

The fact I use the contactless smartphone payments service would make a good impression on other people.

•

The use of the contactless smartphone payments service would give me social approval.

Behavioral Intention •

Assuming I had access to the contactless smartphone payments service, I would intend to use it.

•

Given that I had access to the contactless smartphone payments service, I predict that I would use it.

43

Highlights •

A model integrating perceived risk into a multi-sided perceived value is constructed.

•

Broader view of the duality Gain-Cost in the value perception of an ICT is achieved.

•

Psychological and privacy risks undermine adoption of contactless mobile payments.

•

Value perception of contactless mobile payments is not influenced by media reports.

•

Consumers see more gains than costs in the use of payments with NFC smartphones.

44