What Makes People Actually Embrace or Shun Mobile Payment : A Cross-Culture Study

Mobile payment is becoming increasingly popular, but it encounters the resistance from certain user groups. *is study examines the factors that influence both the technology acceptance and actual usage aspects of mobile payment adoption from the perspective of the general systems theory. Based on a literature review, it conceptualizes the embedding relationships among relevant behavioral processes, personal characteristics, and extrinsic factors and develops a research model. Together, the extrinsic factors in terms of culture, subjective norm, and socioeconomic status and main personal characteristics including demographics, personality traits, and past behavior are hypothesized to have direct and moderating effects on mobile payment acceptance and usage. *e observations collected from China and the USA support most of the hypothesized relationships and reveal interesting cross-culture differences. Whereas users in the USA appear to be more rational and risk-averse, people in China seem more subject to social influence. *e findings contribute to the mobile payment literature by deepening the understanding of adoption stages and expanding the scope of explanatory variables beyond technology acceptance.


Introduction
Mobile payment, by its name, refers to the use of mobile devices and wireless technologies to make payments for goods, services, and bills [1].Along with the fast pace of smartphone diffusion in the recent years, mobile payment has become increasingly popular [2].Yet mobile payment requires mobile data services, and there has been resistance from a large proportion of people due to concerns such as security and privacy [3,4].Due to the inertia, many mobile payment services fail to reach intended customers.
Quite a few studies investigate why people use mobile payment, yet there still exist several research gaps, the most prominent of which concern the overemphasis on technology adoption and the lack of multination analyses [5].Most studies on mobile payment adoption just predict behavioral intention with technology-related perceptions based on theoretical frameworks like the technology acceptance model (TAM) and the unified theory of acceptance and use of technology (UTAUT), but actual usage is what really matters [6].Single-country samples further weaken the generalizability of findings as each market is unique, especially in the cultural influence on user behavior [7].
e main goal of this study, therefore, is to understand why people actually use or avoid mobile payment in distinct cultural contexts.It is essential to include other relevant variables than just technology-related perceptions and collect empirical observations from different countries.As an effort, this study identifies personal characteristics and extrinsic factors pertinent to mobile payment and examines their relationships with both technology acceptance and actual usage.For cross-culture comparisons, it draws samples from China and the USA where people are familiar with mobile payment development but cultures are different.In this way, this study responds to the call for more meaningful research on mobile payment user behavior.e remaining of this article is organized as follows: First, it conducts a literature review that leads to a systems conceptualization of mobile payment adoption.Based on it, a research model is developed to hypothesize the relationships involved in the phenomenon.en, it describes the methodology to collect observations for hypothesis testing and cross-culture comparison.Based on the results, theoretical and practical implications are discussed.

Theoretical Background
Mobile payment is a complex sociotechnical phenomenon that requires a holistic understanding.e general systems theory views such a complex phenomenon as a cohesive conglomeration of interdependent parts for adaptation to the environment [8].From such a perspective, mobile payment adoption can be viewed as a system that comprises behavioral processes, personal characteristics, and extrinsic factors, the interactions among which shape the behavioral tendency of an individual on whether to use mobile payment or not. Figure 1 categorizes the elements involved in such a system as identi ed from the literature.At the center are the behavioral processes of mobile payment adoption, which includes two stages: technology acceptance and actual usage.Behavioral processes are a ected by personal characteristics and extrinsic factors.Speci c to each individual, personal characteristics include demographics, personality traits, and experience/habit associated with the mobile payment.On the other hand, extrinsic factors exert in uences through social mechanisms including culture, social in uence, and socioeconomic status.Depending on their "closeness" to behavioral processes, personal characteristics and extrinsic factors have di erent layers: internal-layer past behavior and socioeconomic status at the bottom, midlayer personality traits and social in uence in the middle, and external-layer demographics and culture at the top.
Technology acceptance is the most studied aspect of mobile payment research.
e main predicting variables including perceived ease-of-use and perceived usefulness from TAM and similar constructs like e ort expectancy and performance expectancy from UTAUT are used by the majority of mobile payment adoption studies [6].Yet technology acceptance is just a necessary condition of actual usage, which has rarely been included in empirical analyses.
e original TAM included the path from behavioral intention to usage behavior [9], yet most studies on mobile payment adoption based on TAM and related frameworks stop at behavioral intention.Actual usage of mobile payment is more than just a yes-or-no decision as typically operationalized for traditional system use.Rather, it concerns how much money, to which extent, and for what purposes a person makes payments with the use of mobile technologies.
Regulating the behavioral processes of mobile payment adoption are personal characteristics of users.Among all, gender, age, personal innovativeness, risk aversion, experience, and habit are identi ed as the most relevant to mobile payment [5].Gender and age describe user demographics that concern not only mobile payment but also general information technology adoption [10,11].On the other hand, personal innovativeness and risk aversion pertain to the personality traits that are closely related to mobile technology adoption and nancial transactions [12,13].Mobile payment is a technological innovation that dramatically changes the way in how people make nancial transactions, especially in the countries where cash is still the main method, and brings some uncertainties.us personal innovativeness and risk aversion represent two sides of a coin related to mobile payment adoption [14].Finally, both experience and habit are related past behavior that inuences future usage of general information technologies as well as mobile payment in speci c [1,15].
Compared with personal characteristics, extrinsic factors are less studied in the extant mobile payment literature.In particular, culture has rarely been included in adoption studies, but it is supposed to have an impact on user behavior.On the other hand, social in uence has drawn more attention, as it and the similar construct of the subjective norm are frequently included in the general information technology adoption studies based on TAM and UTAUT [11].Meanwhile, the variables associated with socioeconomic status including income and education are occasionally included but still underrepresented in mobile payment adoption studies [6].

Research Model
e systems view of mobile payment adoption leads to the development of a research model to examine the relationships among relevant variables.As shown in Figure 2, the model includes two stages of behavioral processes: technology acceptance on the left side and actual usage on the right side.e rst stage's core comprises TAM constructs, which are a ected by personality traits and social in uence.Compared with other re ective psychological constructs in the model, the nal dependent variable in the second stage is a formative construct to capture actual usage with mobile payment frequency, scope, and amount.User demographics, socioeconomic status, past behavior, and culture play different moderating roles.
Compared with traditional payment methods, mobile payment enables users to make nancial transactions anywhere and anytime with great convenience [16].From the perspective of TAM [9], individual perception of the technology in terms of perceived usefulness and perceived   Mobile Information Systems ease-of-use directly affect the intention to use it.As for mobile payment, it is found that such user cognitions have direct impacts on behavioral intention [17].Moreover, existing studies also found that users' evaluation of ease-ofuse positively affects their belief in usefulness for general information technologies as well as mobile payment [18].
H1: Perceived Usefulness has a positive linear relationship with Behavioral Intention.
H2a: Perceived Ease-of-Use has a positive linear relationship with Behavioral Intention.
H2b: Perceived Ease-of-Use has a positive linear relationship with Perceived Usefulness.
As a personality trait, personal innovativeness describes the tendency of an individual to try out new technologies and innovations [19].Empirical evidence suggests that personal innovativeness affects user acceptance of IT-based innovation [20].Mobile payment is an IT-based innovation in mobile commerce, and personal innovativeness has an impact on user evaluation of technology usability [16,21].Highly innovative users are likely to have a more positive attitude towards new technologies in terms of the desire to acquire new skills than less innovative users [17,22].For mobile payment users, therefore, personal innovativeness may have a partial mediating relationship with behavioral intention through perceived ease-of-use.
H3a: Personal Innovativeness has a positive linear relationship with Behavioral Intention.H3b: Personal Innovativeness has a positive linear relationship with Perceived Ease-of-Use.
On the other side of the coin to personal innovativeness, risk aversion is a personality trait that has a negative implication for information technology adoption [23,24].Perceived risk is the manifestation of risk aversion pertaining to the use of specific technologies/innovations that may expose individuals to certain loss or harm [25].For mobile payment, in particular, perceived risk is the biggest concern that prevents users from accepting the new technology [26].Mobile payment users mainly worry about unauthorized use, concerns on device and network reliability, privacy leaks, and transactions errors [27].When people are aware of the potential loss or harm from the use of a system, they tend to downgrade its value and usefulness and hesitate to use it [4,16].H4a: Perceived Risk has a negative linear relationship with Behavioral Intention.H4b: Perceived Risk has a negative linear relationship with Perceived Usefulness.
Subjective norm captures the social influence on the use of a new system from the relevant views and actions of the peers who have direct or indirect experiences with it [28].Compared with traditional methods, mobile payment brings obvious advantages as well as potential risks.Facing the dilemma, an individual usually observes the behavior of surrounding people and seeks advice from peers to get more convinced [21,29].erefore, the subjective norm is found to affect people's willingness to use mobile payment [16].e technology not only supports consumer-business transactions but also integrates seamlessly with social media for personal transfer (e.g., digital "hongbao" or red envelope).e more the people around use mobile payment, the more likely a person is to perceive its value due to network externality [12,27].Meanwhile, others' positive view and active use of mobile payment may mitigate the individual's fear of uncertainty.

Mobile Information Systems
H5a: Subjective Norm has a positive linear relationship with Behavioral Intention.
H5b: Subjective Norm has a positive linear relationship with Perceived Usefulness.
H5c: Subjective Norm has a negative linear relationship with Perceived Risk.
Behavioral intention is widely regarded as the antecedent to actual technology usage at the individual level [30].Yet actual usage is rarely included in the empirical analyses of mobile payment adoption.One study on mobile wallet adoption found that behavioral intention explained a large portion of the variance in usage behavior, the operationalization of which was oversimplified though [10].More clearly defined and measured, actual usage will be included in this study to test its relationship with behavioral intention.H6: Behavioral Intention has a positive linear relationship with Usage Behavior.
Many existing studies based on TAM and UTAUT examine how user demographics moderate the relationships between behavioral intention and its predictors.In this study, personal innovativeness and perceived risk reflect the personality traits that help explain perceived ease-of-use and perceived usefulness.If user demographics serve as the moderators for all of them, their effects are likely to be confounded.Rather, it makes more theoretical and practical sense to investigate the interactions among personal characteristics in terms of user demographics and personality traits.
Men and women vary in their overall attitude toward computers and associated usage behavior [31].Two genders exhibit different perceptions and behaviors due to their different socially constructed cognitive structures to encode and process information [32].ey have distinct perceptions of innovative technologies: males care more about usefulness and relative advantage of systems [33,34], and females are more concerned about ease-of-use and subjective norm [33][34][35].Gender differences are also noticed in the studies of web-based shopping and mobile banking adoption as women are generally more risk-averse than men [36,37].For user adoption of mobile payment, therefore, gender is likely to interact with personal innovativeness and perceived risk on their effects on behavioral intention.H7a: Gender moderates the relationship between Personal Innovativeness and Behavioral Intention.
H7b: Gender moderates the relationship between Perceived Risk and Behavioral Intention.
User perceptions and attitudes toward computer technologies also vary across age groups [11].Mobile payment involves smartphone usage, and the learning curve becomes steeper when age increases.us, age is found to moderate the effects of effort expectancy and social influence on user intention in mobile learning [38].As an innovation involving financial transactions, mobile payment is likely to follow a similar pattern of adoption to that of online shopping, which is also subject to age disparity [39,40].Compared with young adults, seniors experience more barriers to online shopping due to risks and habits [41].Similar to gender, age is likely to play a moderating role.
H8a: Age moderates the relationship between Personal Innovativeness and Behavioral Intention.H8b: Age moderates the relationship between Perceived Risk and Behavioral Intention.
Computer technologies require users to have certain knowledge and skills, and their education levels make a difference in adoption and usage behaviors [42].How well a user is educated is associated with the person's perception of a system in terms of its usability [43].Also, education level is found to be negatively correlated with user anxiety in computer use [44].For an innovative technology like mobile payment, therefore, education may interact with personality traits related to innovativeness and risk-averseness.
H9a: Education moderates the relationship between Personal Innovativeness and Behavioral Intention.H9b: Education moderates the relationship between Perceived Risk and Behavioral Intention.
Whether mobile payment is for online shopping or face-to-face purchase (e.g., restaurant and taxi), it is the last step to complete the transaction.Due to other constraints, a person's intention to use mobile payment may not always be converted into actual usage.Among them, the individual's previous mobile payment experience and income level cannot be ignored [45].On the technological side, the previous experience with an innovation can influence an individual's perceived ease-of-use, which affects usage volume and frequency [43].On the socioeconomic side, personal income is closely related to purchasing power and risk tolerance associated with online transactions [46].In a study of mobile wallet, for instance, income is found positively associated with an individual's acceptance and use of the technology [10].All else being equal (especially intention), users at the different experience and income levels are likely to use the mobile payment to different extents.H10: Experience moderates the relationship between Behavioral Intention and Usage Behavior.H11: Income moderates the relationship between Behavioral Intention and Usage Behavior.
Finally, mobile payment platforms are based on specific currency systems, and people's usage behavior is likely to vary from one country to another.In particular, national culture concerns the fundamental values and shared beliefs among people in a country [47].ere are six cultural dimensions along which people's behavior may vary: (a) Power Distance: acceptance to unequal power distribution in society; (b) Individualism versus Collectivism: tendency integrate into strong cohesive groups; (c) Masculinity versus Femininity: preference between male-associated qualities (e.g., assertiveness and material success) and femaleassociated ones (e.g., modesty and quality of life); (d) Uncertainty Avoidance: fear of unknown situations; (e) Longterm Orientation: persistence and thrift leading to future 4 Mobile Information Systems rewards; (f ) Indulgence: tendency to seek happiness [48].Among them, some are closely related to the extrinsic factors and personal characteristics pertaining to mobile payment.
In particular, the dimension of Individualism versus Collectivism concerns social influence and the dimension of Uncertainty Avoidance concerns risk aversion.us, the hypothesized relationships as mentioned above may vary significantly across different cultures.

Research Design.
e target population comprises mobile payment users in multiple countries that have relatively high population penetration of smartphone technology yet very different cultures.e two countries that lead the trend of smartphone diffusion are the USA and China [49].Whereas mobile payment in the USA had an early start, China is catching up quickly, and the total annual transaction exceeded 5 trillion US dollars (50 times that of the USA and more than Japan's GDP) in 2016 [50].eir cultures are distinct, as shown in Figure 3.In particular, the USA is high in individualism whereas China is high in collectivism.is suggests that the effects of social influence on mobile payment user behavior vary across two countries.Also, the USA is noticeably higher than China in uncertain avoidance, which makes a difference in perceived risk associated with the mobile payment.
is study tests the invariance in the hypothesized relationships with the observations collected from different cultures.If a large proportion of the relationships vary significantly across the samples, there is supporting evidence of the cultural influences on mobile payment adoption.
us, this study conducted a survey with working professionals in both China and the USA, most of whom own smartphones and are more likely to use mobile payment than students and retirees.Invitations to the online survey were sent to full-time workers in two countries based on snowball sampling.Initial contacts were gathered from three profession training programs in China and USA.
e participants were encouraged to send the invitation to their friends and relatives who might have used mobile payment.

Subjects.
Altogether, there were 162 valid responses in the China sample and 136 in the USA sample, leading to a total sample size of 298.e sample size is sufficient for statistical analyses used in this study, mainly factor analysis for measurement validation and partial least squares for model estimation [51].As shown in Table 1, the participants in the two samples had somewhat different profiles.Whereas more than one-third participants in the USA sample had used mobile payment for three years or more, one-fourth had limited experience (i.e., <6 months).Meanwhile, very few in the China sample were new to the technology but more than 80% were quite experienced with more than oneyear history.On average, the China sample was younger, and the USA sample had higher levels of education.Gender and income distribution were relatively balanced between the two samples (e.g., around two-thirds in both had low or medium low income).
Potential nonresponse bias was assessed by comparing early and late responses [52].In both samples, the first 45 and last 45 responses had insignificant differences on the means of any variables based on t tests.e invariance suggested no serious threat of nonresponse bias.

Measurement.
e Appendix section lists all the measurement items used in the questionnaire.All psychometric scales were adapted from previous studies.Items measuring Perceived Usefulness and Perceived Ease-of-Use were adapted from Davis [9].Behavioral Intention and Subject Norm measures were adapted from Fassnacht and Köse [53] and Schierz et al. [54].Personal Innovativeness and Perceived Risk were measured with items adapted from Yang et al. [16].Other more objective variables, such as mobile payment frequency, scope, and amount, were measured with self-developed items.e score of mobile payment scope was calculated as the count of total options (e.g., dining and bill pay) checked.
As most of the questionnaire items were psychometric measures, their potential common method bias was assessed following Harman's one-factor test [55,56].Exploratory factor analysis (EFA) results revealed that 40.11% common variance was captured by the first principal component (less than half), whereas all the major components (eigenvalue >1) explained 69.23% (more than two-third).Confirmatory factor analysis (CFA) results as reported in Table 2 compared method-only, trait-only, and trait/method models.e goodness-of-fit indices of the method-only model were much worse than those of trait-only model, which were even slightly better than those of trait-and-method model.Together, the EFA and CFA results indicated that common method bias was not a serious issue.

Results
e descriptive statistics in Table 3 show the response patterns of all variables, the possible range of each is between one and five.In the overall sample, the average responses of psychological variables related to technology acceptance were higher than the midpoint of three, but those of actual mobile payment usage variables were lower.
e distinct response patterns support the use of relatively objective Usage Behavior measures that not only mitigates common method bias but also gauges the gap between psychological behavior and overt behavior.Cross-country comparison shows that the USA sample exhibited more positive responses on psychological constructs (except for Perceived Risk), but the China sample showed relatively active mobile payment usage.
e results in Table 4 validate the reflective psychological constructs in terms of convergent validity, discriminant validity, and nomological validity.Convergent validity was supported as all the coefficient alpha and composite reliability (CR) values were above 0.7, and average variance extracted (AVE) values were above 0.5.Discriminant Mobile Information Systems validity was supported as the square roots of AVE were all greater than the correlation coe cients.As expected, all variables were positively correlated with each other, except for perceived risk.us, nomological validity was also supported.
e validation of Usage Behavior as a formative construct has di erent requirements.Instead of being consistent with each other, formative indicators are supposed to be somewhat distinct and have nontrivial contributions to the construct in question.As shown in Table 5, all the variance in ation factors (VIF) were well below 5, indicating nonsalient multicollinearity among formative indicators.e relationship between each indicator and the construct was signi cant as indicated by multiple regression weight, and  To test the hypothesized relationships that involve both reflective and formative constructs, partial least squares (PLS) structural equation modeling is appropriate [51].Table 6 reports the endogenous variables' coefficients of determination (R 2 ) for the overall sample as well as two country samples.In the overall sample, more than twothirds of variance was explained for Behavioral Intention, less than half for Perceived Usefulness and Usage Behavior, around one-fourth for perceived ease-of-use, and almost none for Perceived Risk. is is somewhat consistent with the number of predictors that each construct has.In particular, the majority of variation in Behavioral Intention was accounted for, suggesting that most important predictors are included.Across the USA and China samples, the coefficients of determination from the former were more or less higher than those from the latter, especially in the case of Usage Behavior.us the gap between mobile payment intention and actual usage seems wider for people in China than those in the USA.
Table 7 reports the standardized estimates of each path coefficient obtained from overall and split samples.In the overall sample, all the hypothesized linear relationships (i.e., H1-H6) were significant except for that between Subjective Norm and Perceived Risk.Meanwhile, three out of eight moderating effects turned out to be significant, including the two from Experience and Income to the relationship between Behavioral Intention and Usage Behavior.In either country sample, however, only the moderator Education did not yield any significant effects.us, there is more or less supporting evidence for each research hypothesis, except for those related to Education.In addition, a multigroup analysis (MGA) was conducted to examine cross-culture differences in path coefficient estimates.e observed significance level of each difference was obtained with the permutation method of MGA based on the two-tailed test [51].About half of the relationships were found to be quite different across the USA and China samples.us, culture did make noticeable differences in hypothesized relationships.
In particular, Perceived Usefulness and Perceived Easeof-Use had stronger effects on Behavioral Intention in the USA sample than in the China sample.Personal innovativeness' effect on Behavioral Intention, however, was the other way around.Perceived Risk, as another user characteristic, had more negative effect on Perceived Usefulness in the USA sample than in the China sample.Social influence (i.e., Subjective Norm) on Behavioral Intention was stronger in China than in the USA, yet its effect on Perceived Risk switched in strength between two samples.Gender interacted with Personal Innovativeness in China but with Perceived Risk in the USA in their effects on Behavioral Intention.Age is the opposite: it interacted with Perceived Risk in China but with Personal Innovativeness in the USA.Finally, Experience and Income played more negative moderating roles on the relationship between Behavioral Intention and Usage Behavior in the China sample than in the USA sample.eir direct impacts on Usage Behavior were also more positive in the USA sample.
Figure 4 illustrates the salient moderating effects in each country sample.eir f-square values indicate the effect sizes of moderation [57].All were well above 0.009, the average moderating effect found in a meta-analysis [58].When the age increased, the effect of perceived risk on behavior intention got less negative in the China sample and that of personal innovativeness got more positive in the USA sample.Compared with males (Gender � 0), females (Gender � 1) saw a more positive relationship between Personal Innovativeness and Behavior Intention in the China sample, but more negative effect of Perceived Risk in the USA sample.For both countries, more Experience in mobile payment meant more active Usage Behavior, yet the effect of behavioral intention diminished due to habitual use   Mobile Information Systems 7 [15].Income played different moderating roles across two countries: people with higher income were more likely to convert Behavioral Intention to Usage Behavior in the USA sample, but it was the opposite in the China sample.

Discussions
e findings yield some important theoretical and practical implications.First of all, they support the conceptualization of mobile payment adoption as a system that involves the interactions among behavioral processes, personal characteristics, and extrinsic factors.ree layers of personal characteristics and extrinsic factors affect two stages of behavioral processes in terms of technology acceptance and actual usage in different ways.e internal-layer elements at the bottom of extrinsic factors and personal characteristics in Figure 1, including past behavior and socioeconomic status, are found to mainly moderate the relationship between technology acceptance and actual usage.e midlayer elements in the middle, including personality traits and social influence, have direct impacts on technology acceptance.
e external-layer elements at the top, including demographics and culture, mainly make differences in the linear relationships involved in technology acceptance and other moderating relationships.
e multistage and multilayer conceptualization and modeling yield a deeper understanding of mobile payment adoption.Compared with extant research on mobile payment adoption, this study helps bridge the gap between the psychological behavior of technology acceptance and the overt behavior of actual usage with additional variables associated with both.Due to their different natures, personal characteristics and extrinsic factors exert influences on technology acceptance and actual usage through direct, mediating, and moderating routes.In addition, the comparison between the samples from China and the USA suggests that their cultures make differences in the strengths of many relationships.Responding to the call for more meaningful and generalizable research on mobile payment adoption [5], this study contributes to the mobile payment literature by deepening the understanding of adoption stages and expanding the scope of explanatory variables at the same time.
Some specific findings may be interesting to researchers and practitioners.For instance, the effect of Subjective Norm on Behavioral Intention was significantly stronger in the China sample than that in the USA sample.
e two countries are very different along the relevant cultural dimension of Individualism versus Collectivism.Compared with American people, Chinese people are more likely to form strong and coherent groups and influence each other.On the other hand, the negative effect of Perceived Risk on Perceived Usefulness was stronger in the USA sample than that in the China sample.Along the relevant dimension of Uncertainty Avoidance, correspondingly, the finding suggests that American users worry more about the potential security and privacy breaches from using mobile payment than Chinese users.
In addition to the moderation of linear relationships, Culture makes a difference in how Income moderates the relationship between Behavioral Intention and Usage Behavior.In the USA sample, higher income is conducive to the conversion from technology acceptance to actual usage, but it is the opposite in the China sample.In China, credit  Mobile Information Systems card transactions are still rare and the primary method of payment is still cash.Mobile payment provides a viable means to shop online, dine in restaurants, and pay for services (e.g., taxi).Online stores usually o er lower prices than brick-and-mortar stores.To encourage the use of mobile payment (e.g., so as to keep track of customers), vendors often o er additional discounts.For people of relatively low income, the saving from mobile payment constitutes a major incentive.Yet mobile payment is riskier than cash transactions.For people with relatively high income, they are more concerned about the potential loss associated with security and privacy breaches than monetary saving.In the USA, however, the di erences in prices and risk levels between mobile payment and other methods are not that obvious, and income plays a positive moderating role as expected.
Insigni cant results also deserve a close look.Especially, Education does not moderate the e ects of Personal Innovativeness and Perceived Risk on Behavioral Intention in either country.In theory, education may provide potential mobile payment users more background knowledge about it and facilitate adoption decision-making [59].Yet there is no supporting evidence from the observations.One explanation is that mobile payment is an innovation that brings people not only bene ts but also risks.When people are well-educated, they also become aware of its cons as well as pros, canceling out the positive e ect of Education.is explanation is more applicable to developing countries like China where mobile payment is still relatively new to most people.In developed countries like USA, however, mobile payment is no longer a cutting-edge innovation, and it is possible that people at all education levels are familiar with it.e findings provide some helpful clues on the best practices to promote mobile payment adoption.ere are two possible routes: one to enhance technology acceptance and another to materialize actual usage.At an early stage of mobile payment development, it is worth the effort to help people accept the technology first.It is more effective to target potential users of relatively high personal innovativeness and low risk aversion, who will then influence others through word-of-mouth.After mobile payment gained a certain level of popularity, the main challenge is how to convert technology acceptance financial transactionmaking.
is highlights the importance of investigating actual usage in addition to behavioral intention at the current stage of mobile payment diffusion.As the findings indicate, the strategy needs to be based on case-by-case analyses considering cultural factors, business environment, income levels, and so on.

Conclusion
is study examines how personal characteristics and extrinsic factors influence the behavioral processes of mobile payment adoption in terms of technology acceptance and actual usage.Based on the understanding of embedding relationships, it proposes a research model that hypothesizes their direct and moderating effects on user behavior.e survey observations collected from the USA and China provide supporting evidence to most of the research hypotheses and reveal some interesting cross-country differences.
e findings suggest that social influence and personality traits have direct impacts on technology acceptance, whereas demographics, past behavior, socioeconomic status, and culture play different moderating roles.
is study has limitations that point to the directions of future research.A major limitation of this study is due to the fact that the observations were collected from only two countries.China and the USA are selected because they have large populations of smartphone users and are distinct in culture.Yet they cannot represent other countries and regions, which limits the generalizability of findings.In addition, culture is used as a grouping variable in this study, but its different dimensions may have different effects on mobile payment adoption.Future studies may collect data from more countries and include specific cultural dimensions in analyses.is will not only enhance the generalizability of findings but also reveal the specific roles that different cultural dimensions play.

Perceived Usefulness
Using mobile payment makes my life more convenient.Compared with other methods, mobile payment has many advantages.To me, mobile payment is useful.
Perceived Ease-Of-Use It is easy for me to become skillful at using mobile payment.
e steps to follow for mobile payment are clear to me.In general, I find mobile payment easy to use.

Figure 1 :
Figure 1: A systems view of mobile payment adoption.

Table 2 :
Common method bias assessment with CFA.
Note.RMSEA: root mean square error of approximation; CFI: comparative t index; NFI: normed t index.

Table 3 :
Descriptive statistics and sample comparison.

Table 1 :
Pro les of participants.

Table 4 :
Measurement validation for reflective constructs.Cronbach's coefficient alpha; CR: composite reliability; AVE: average variance extracted; ns not significant at 0.05 level, all other correlation coefficients were significant at 0.01 level.e bold on the diagonal of correlation matrix indicates the squared root of AVE.

Table 5 :
Validation of usage behavior as a formative construct.
Note.VIF: variance inflation factor.All weights and outer loadings were significant at 0.01 level.

Table 6 :
Coefficients of determination (R 2 ) of endogenous variables.

Table 7 :
Standardized PLS estimates.significant at 0.1 level; * * significant at 0.05 level; * * * significant at 0.01 level. * Behavioral IntentionI intend to use mobile payment.I will use mobile payment if there is a chance.I will recommend mobile payment to my friends/ relatives/colleagues.