India is the second-largest mobile market after China with the population of 1.3 billion people. The number of smartphone users in India has increased drastically. As per Ericsson Mobility Report, India grew the most in terms of net additions during the first quarter of 2017 (+43 million) and that the mobile subscriptions are expected to reach 1.4 billion by 2022. (Ericsson, June 2017). The fierce competition among the mobile phone makers and the entry of 4G technology causing lowering of tariff plans have made the purchase and use of smartphones more affordable thereby increased the penetration of smartphones further in recent months.
The government is promoting a program known as the Aadhaar initiative which assigns a unique identification number to every registered citizen based on fingerprints and IRIS data— similar to a U.S. Social Security number. This biometric information allows people to access government services more easily, such as the use of UPI based apps, government subsidies, health care and obtaining government certificates, or do things like opening a bank account remotely using smartphones. The government is pushing the use of Aadhaar based identification and use of devices like smartphones to further its cause of transparency, and corruption-free society. The available modes of cashless transactions in India are Banking Cards (credit, debit cards), USSD (Unstructured Supplementary Service Data), AEPS (Aadhaar enabled payment system), UPI, Mobile Wallets, Banks Pre-paid Cards, Point of Sale (handheld device with card and /or biometric reader), Internet Banking, Mobile Banking and Micro ATMs (GOI, 2017).
Spurred by the demonetization and the subsequent increase in the daily transaction limit of digital wallets to Rs 20,000/-, India has seen a phenomenal increase in the number of digital wallets. Of late various digital wallets start-ups have sprouted up like Airtel Money, Citi MasterPass, Citrus Pay, Ezetap, Freecharge, HDFC PayZapp, ICICI Pockets, JioMoney, Juspay, LIME, Mobikwik, MomoeXpress, MoneyonMobile, Mswipe, Ola Money, Oxigen, PayMate, Paytm, PayUmoney and State Bank Buddy to name a few leading ones. (sumHR, 2017)
Demonetisation, launch of UPI, linking of financial information and mobile numbers with biometric identification has facilitated the use of smartphones for e-payments. However, the technology-related apprehensions, inertia to change and other issues have seized the pace of adoption of mobile-based e-payments by the masses. This study is limited to the e-payment method affected by the use of smartphones and the impact of factors associated with the adoption of e-payment on smartphone users.
There is quite a good deal of empirical studies available, suggesting several factors affecting the adoption of technology which has relevance to the current study concerning the adoption of an electronic mode of payment by the smartphone users. Perceived mobility, perceived trust, ease of use, environmental risk, perceived utility, perceived reputation, attitude etc have been cited by many as some of the factors which mingle up with the adoption of technology.
The innovation diffusion theory (IDT) has been one of the strong theories to predict the diffusion of innovations in a social system. The innovation diffusion theory (IDT) describes how innovations or technology becomes accepted and spread through societies (Rogers, 2003). The innovation-decision process In IDT is a five-stage process comprising of (Knowledge stage, Persuasion, Decision, implementation and Confirmation), through which a person acquires knowledge about the innovation, forms an attitude about the innovation, and then makes a decision to accept or reject the innovation, if a the decision is favourable, the person implements the innovation and proceeds to confirm their decision. Here, the specific attributes of a given innovation, such as relative advantage, compatibility, complexity, trialability, and observability govern the rate of adoption (Rogers, 2003). Additionally, there are four variables that may further influence the diffusion like the type of innovation, the communications channel used, the nature of the social system such as a degree of interconnectedness, and norms and the fourth is the extent of the change agents’ promotion efforts (Rogers, 2003). IDT also provides some structure for evaluating individuals as Innovators, early adopters, early majority, late majority and laggards (Rogers, 2003).
However, this theory does not provide evidence of how innovation characteristics fit into the decision process and how decisions from attitudes (Karahanna et al., 1999) also the theory does not explain fully, how the attributes of the innovation help in forming attitudes.
The other extensively referred, Technology acceptance model (TAM) was introduced by Davis in his Doctoral thesis to explain the factors affecting the individual’s acceptance of an information system (Davis, 1989). According to this model, the external observable variables influence the two latent variables perceived ease of use and perceived usefulness which in turn develops individual’s predisposition towards technology. This change in attitude brings about expected behavioral intention to use the technology and determines the final status of adoption of the technology by the concerned individual.
Technology acceptance model (TAM) itself is based on the theory of reasoned action (TRA) developed by Martin Fishbein and Icek Ajzen in 1967, which suggests that the individual behavior is guided by behavioral objectives that are the function of an individual’s attitude toward the behavior and subjective norms surrounding the performance of the behavior. This model forms the backbone of studies associate with attitude-behavior relationships.
Taylor and Todd (1995) suggested a modification in TAM by integrating the theory of planned behavior called TAM?TPB model. The theory of planned behavior is an extension of the theory of reasoned action (Ajzen, 1991). The Theory of Planned Behavior (TPB) was developed to predict behaviors in which individuals have incomplete volitional control (Ajzen, 1991).
In 1992 Davis et al. came up with Motivational Model additionally recognizing the perception of pleasure and satisfaction as intrinsic motivational factors. Later in 2000 Davis along with Venkatesh proposed Extended TAM2 Model which included social influence processes and cognitive instrumental processes as well. In 2003 Venkatesh et al. put forward the Unified theory of acceptance and use of technology (UTAUT) with performance, effort, social influence and facilitating conditions as factors.
In a study about smartphone adoption Tahir Ahmad Wani and Syed Wajid Ali have underscored the role of Perceived Characteristics of the innovation (like Relative advantage, Compatibility, Complexibility, trialability and observability) as well as The intrinsic factors of motivation represented by Perceived Employment, in technology adoption process these where moderated by characteristics of decision making unit and communication channel in their model (Tahir Ahmad Wani, 2015). They have also suggested the inclusion of constructs like perceived enjoyment and risk in TAN model (Tahir Ahmad Wani, 2015).
The construct, Self-efficacy was also used in some studies and has been shown as positive relationship with behavioral intentions as in (Jengchung V. Chen, 2010)
In a nutshell, innumerable model variations and approaches have been reported by several researchers and that the models have been used in varied scenarios see (Surendran, 2012 ; Kumar, 2013; Alaa M. Momani, 2017; G.D.M.N. Samaradiwakara, 2014; Rajesh Sharma, 2014). The researchers have also reported varied associations and deviations in the behavior of usual variables from expected norms/models for instance (?kram Da?tan, 2016). They did not find any association of perceived ease of use and perceived usefulness with attitude. It can, therefore, be argued that even with the widely used constructs and replication of existing model, one does not seem to add any further knowledge about the subject. For that reason, the approach adopted in this study is to start afresh with focus group interactions and find out factors which concern the most for the subjects in the adoption of e-payment system.
This study shall comprise of two parts, Firstly, an exploratory study will be conducted via focused group interview method to understand and enlist all the possible concerns of the subjects (smartphone bearing respondents) which could potentially have a bearing in the adoption of e-payment. These concerns would then be assimilated into a set of observable constructs for development of questionnaire thereafter the questionnaire would be administered in the second and final phase of the study to carry out an empirical study so that the effect of extracted principal factors (latent) can be analysed on the adoption of e-payment system by the smartphone bearing respondents.
Principal exploratory factor analysis with varimax rotation would be employed to extract principal factors out of observable constructs employed in the questionnaire to suggest a model comprising of latent variables (extracted factors) and the outcome variable. Factors with Eigenvalues greater than 1 out of all the extracted factors would only be employed in regression analysis (Sharma, 1996); (Tabachnick and Fidell, 2001; Johnson and Wichern, 2002) as in (E. Sakar, 2011). Since within factor loadings are not correlated, the factor loading of these extracted factors would then be employed in logistic regression to predict the outcome variable (status of use of e-payment system) and assess the utility of the model thus created.
Data analysis and findings
In the focus group study, 15 people having varied background revealed some of the prominent factors which could potentially impact the adoption of the e-payment system by the people. These issues exhibited by the people were identified and isolated as mentioned in Table–1. A questionnaire was developed from the issues mentioned in the table–1 bearing the prominently figuring issues. In this pool of possible variables, two more variables, namely, Attitude and Behavioural intentions were purposefully added to gauge the path and relative impact of other mentioned variables on adoption of e-payment system on mobile phones. Subjects were then approached, online (via https://esurv.org/) to collate the responses. A total of 132 proper responses were utilized for the data analysis. SPSS version 23 was used for the data analysis.
Insert Table-1 about list of factors identified for the study
Kaiser-Meyer-Olkin measure of sampling adequacy was 0.887. Usually, a value of more than 0.7 is suggestive of a good sample size for factor analysis. The probability of occurrence of Bartlett’s test of sphericity with the Chi-Square value of 1582.331 was less than 0.001. Therefore the sample was assumed to be adequate to proceed with Exploratory Factor Analysis.
Exploratory Factor analysis with Varimax rotation and Kaiser Normalization yielded 2 principal factors out of thirteen observable constructs put to analysis. These 2 principal factors were selected because they had Eigenvalues above 1. The explained variance in total data out of these 2 Factors, i.e. the extracted (as well as rotation) sums of squared loadings was at 71.57 percent.
Insert Table-2 about Factor analysis – Component wise variance
Interpretation of the extracted factors
Factor-1 appears to be too complex to be explained reasonably because most of the observable variables of the likes of Pre-understanding, Ut-easy-manage, Us-easy-learn, Us-easy-use, Soc-urge, Fu-future-cash, Att-good-Idea, and Afford-smartphone have rounded off factor loadings of 0.7 or more on it. (See description of these abbreviations in table-1 and table-2 for factor loadings).
The Factor-2, however, has only three observable variables that have significant loadings on it. It has a very high loading from Ri-riskiness (-0.952). So this factor can be conveniently assumed to be influenced much by the element of risk. But, since risk has negative loadings onto it, the factor should well be explained by an attribute signifying opposite of Riskiness. Also, the element of trust (Tr_trustworthy) has the significant and positive impact on factor-2 (0.686). This factor calls for renaming it around opposite of Risk and towards trust. Factor-2 also has the negative impact of Lethargy (Inr_lethargy)(-0.758). The positive of which could be like vigor. So this Factor can be a combination of trust and spirits and thus be appropriately renamed as ‘conviction’ or ‘conviction with enthusiasm’.
Insert Table-3 about Factor analysis – Rotated Component Matrix
The purpose of this study was to suggest a model for predicting the probability of adoption of e-payment system by the users of smartphones. Therefore the variable measuring the status of use of e-payment (a dichotomous dummy variable assuming only two values 0 for nonusers and 1 for users of e-payment system) was regressed on the factor scores to assess the predictive powers of the factor scores.
The logistic regression with the above mentioned 2 factors as predictors revealed that the predictability of outcome improves from 67.4% to 81.4 % percent when the intercept-only model is compared with model having both the factors. The logistic regression model was statistically significant, ?2= 41.729, p