E-Trust Integration with Technology Acceptance Model (TAM) in Adoption of Digital Financial Services (E-Cash)

This research is shown to know how the trust (e-trust) integrated with the technology acceptance model can affect the process of digital financial services digital service adoption. The results of this study indicate that 1) Perceived Ease of Use positive and significant effect on Perceived of Usefulness, 2) Perceived Ease of Use positive and significant effect on Attitude Towards Using, 3) Perceived of Usefulness positive and significant effect on Attitude Towards Using, 4 ) Perceived of Usefulness positive and significant effect on Behavioral Intention to Use, 5) Attitude Towards Using positive and significant effect on Behavioral Intention to Use, 6) E-trust positive and significant effect on Perceived of Usefulness, 7) Positive E-trust and significant effect on Attitude Towards Using, and 8) E-trust positive and significant effect on Behavioral Intention to Use to the process of re-adoption of digital financial services. Limitations in this study, one of which is the dimension of trust that is used only limited to the dimensions of credibility, benevolence, integrity and orientation to resolve problems.

(2016) to be applied in the process of adoption of digital financial services or digital financial services as one of the banking products that is being promoted to reduce the use of cash in Indonesia.

II. LITERATURE REVIEW
Consumers are assumed to go through five stages to arrive at the decision to buy or reject a new product. These five stages can be described as follows: 1) Awareness stage. In this first stage, consumers are exposed to a new product. They become aware and recognize the existence of the product. 2) Stage of interest or knowledge. When consumers are aware of the presence of a new product, they begin to attract and seek information to find out more about the product. 3) Evaluation stage. Based on information obtained in the previous stage, consumers will conduct an evaluation, and if the results are positive, consumers will try it. 4) Stage of trying. Consumers try or buy in small quantities. Experience in this stage will determine whether consumers will adopt or reject the product. 5) Step adoption (or reject). At this stage the consumer decides to use the product and buys in the required amount, or refuses to use it. Consumers will go through the stages of the decision process before deciding to buy, use or consume a product / service. Schiffman and Kanuk (2007: 8) describe consumer decision making through simple models that are divided into three stages, namely the input stage, process and output. Stage input is an external factor that affects the introduction of consumers to the needs of a product. External factors consist of two sources, namely: first, the marketing efforts undertaken by the company to influence the consumer, namely the product itself, its price, its distribution and its promotion, and secondly: from the sociological influence of the consumers, ie family, friends, neighbors, other informal sources, social class, culture, and sub-culture. The influence of the two sources is an input for consumers who determine what consumers will buy and how they use what they have bought (Suprapti, 2010: 16). Davis (1989) proposed a theory of Technology Acceptance Model (TAM) that explains how computer adoption by technology-based beliefs (usefulness and ease of use ). TAM theory is backed by two theoretical branches namely Theory of Reasoned Action (TRA) (Fishbein and Ajzen, 1975) and Theory of Planned Behavior (TPB) (Ajzen, 1991). TRA describes the beliefs of a user / consumer will determine the attitude and intense, which will further affect the actual behaviors.
Perception of expediency (perceived usefulness) refers to the degree to which a person believes that using a particular system would help improve performance in the job (Davis, 1989). Adamson and Shine (Pardyanto, 2013) define the perception of benefits as constructing one's belief that the use of a particular technology will improve their performance. From these two definitions, it can be concluded that the perception of benefits relates to the productivity and effectiveness of the system of utility in the overall task of improving the performance of the people using the system. Vankatesh states that there is an important influence of benefits in understanding the individual response in information technology.
Attitudes to use can be defined as positive or negative feelings of users of technological systems arising from the effects of ease of use and perceived use benefits. Attitudes will affect the behavior of users in the intention to use and also the use of information systems and technology.
Furthermore, behavioral intention to use is a behavioral tendency to keep using a technology. The degree of use of a technology to a person can be predicted from his attentiveness to the technology, such as the motivation to keep using, as well as the desire to motivate other users. The final variable from Davis (1989) is actual usage, is the real condition of the use of technology, conceptualized in the form of measurement of frequency and duration of time of technology use. Someone will be content to use the system if they believe that the system is easy to use and will improve their productivity, which is reflected in the real conditions of use.
T rust is the belief of a certain party to the other in conducting a transaction based on a belief that the person he believes will fulfill all his obligations well as expected (Rofiq, 2007). Then E-trust in this case can be defined as the consumer's belief in the quality and reliability that the goods or services offer, as well as his beliefs about the trustworthiness, honesty and virtues of e-commerce companies (Sativa and Astuti, 2016).
Digital financial services (LKD) is an activity of payment and finance system services conducted in cooperation with third parties and uses mobile and web-based technology tools and tools in the framework of inclusive finance. In this case LKD in process online, the transaction process is connected directly to the central s I stem host computer LKD to perform authorization and validation prior to commencement of the transaction process so that settlement LKD can be done in real time and / or near real time and available notification of transaction status immediately after a financial transaction.

III.RESEARCH METHODS
The research was conducted in several stages, starting with data collection on the number of users of digital financial services conducted for 34 days starting from 25 January to 27 February 2018, and FGD conducted for 2 days located in the area of Denpasar Sanur. Based on the conceptual model built and the relationship between the variables, the inferential statistics used to process the data are Partial Least Square (PLS) with Smart PLS Version 3.0 program. PLS is a model of Structural Equation Modeling (SEM) equations based on components or variants. PLS almost resembles regression but more than that, it simultaneously combines the Structural Path model (the theoretical relationship between latent variables) as well as measuring the path (the relationship between the latent variable and its indicator).

IV. RESULTS AND DISCUSSION
Profile of respondent's age, it can be seen that digital financial service user respondents consist of several age categories, both respondents aged under 30 years to over 50 years. Most of the compositions are respondents who are under 30 years old. Viewed from the gender, most of the respondents were women i.e as many as 83 people (59.3 percent), and for the marital status of respondents have the same composition.
Profile of respondents viewed from the level of education, grouped into four levels, namely: SMA, Diploma, S1 and S2. From 140 respondents, most of them educated undergraduate. Respondents in the use of digital financial services mostly work as civil servants/BUMN/ POLRI/ TNI, but also there are respondents who work as private employees, students, entrepreneurs and housewives. The majority of highly educated respondents are an indication that their perceptions in answering questionnaires become more objective.
Profile of respondents based on behavior in using digital financial services such as the long-term use of digital financial services, and transactions ever made. Based on the behavior of the use of digital financial services, the respondent's profile is also differentiated based on the length of use in using digital financial services digital services and the type of transaction ever done. Majority users are respondents who use digital financial services for more than 2 years, indicating that respondents are able to better illustrate the perception of the questionnaire so that they can know the reason of respondents in re-adopting the digital financial services in question.
There are three values that must be considered at this stage namely the value of convergent validity, discriminant validity , and composite reliability . Convergent validity is used to find out the instrument items that can be used as indicators of all latent variables. The results of this test are measured by the value of factor loading (outer loading) of the construct indicator. The following test results of convergent validity are presented in Table  1.  Table 1 shows that all outer loading has a value greater than 0.5, so this measurement can be summed up to meet the requirements of convergent validity.
Validity test is also done by testing method comparing the root value of average variance extract (AVE) in each construct with correlation among other constructs contained in the model. Discriminant validity test results are presented in Table 2. which explains that the AVE value in the research variables has values above 0.5 so that these measurements can be explained to meet the requirements of discriminant validity measurement. Furthermore, the composite reliability test is done to test the reliability of the instrument in a research model. Composite reliability test results are presented in Table 3.   Table 3 it can be explained that the results of the composite reliability test are good, since the latent variables have all been reliably where the value of composite reliability is greater than 0.7.This indicates that all indicators have become their own measuring tools. The final step taken after composite reliability testing is testing the value of croanbach's alpha .  Table 4. showed that all latent variables have croanbach's alpha value above 0.7 so it can be concluded that this research has been fulfilled.
Goodness of fit structural in inner model describes relationship between latent variables based on substantive theory. In assessing the model with PLS, it starts by looking at R-square for each endogenous latent variable. Inner model test results can see the relationship between constructs by comparing the significance and R-square values of the research model (Gozhali, 2008: 42).
The R2 value of each endogenous variable in this study can be seen in Table 5. 16. The value of R2 variable intention to use (Behavioral Intention to Use) of 0.777 can be interpreted that 77.7% variation in variable intention to use is explained by the variables used in the model, while the rest of 22.3% is explained by variables or other factors outside the model. 0,777 Goodness of fit on the inner structural model tested the model using values predictif -relevance (Q 2 ) to measure how well the observed values generated by the model and estimation parameters. The predictiverelevance value is obtained by the formula: The above results show a predictive-relevance value of 0.9836, thus the value is> 0. The value of Q 2 approaches value 1 which means the model has a relevant predictive value, and it can be stated that this structural model has fit with the data.
In a research value factor loading (outer loading) also need to be considered, ie shows the weight of an indicator of the variable. The largest factor loading factor explained that the indicator was said to be the dominant variable measure, as shown in Table 6.
As shown in Table 6, the overall variables in the study have a factor value loading more than 0.50, so that means the indicators used are valid to form E-Trust variables, Perceived Ease of Use, Perceived of Usefulness, Attitude, and Behavioral Intention to Use. The mean in the table shows that in general the user's assessment of E-Trust implementation or practice, Perceived Ease of Use, Perceived of Usefulness, Attitude, and Behavioral Intention to Use is good.  Table 6 shows that all indicators of e-trust variables have a factor loading factor above 0.5 with the highest indicator value is in the " credibility " indicator of 0.914 with respondents' perception value of 4.24. This reflects that the competence / expertise, reliability and ability to fulfill the promise that has been formed by the company becomes an important reason for users of a technology in choosing a technology that they will use, thus the credibility of a company becomes one of the strong reasons for users to choose technology which will be used.
When viewed on the perceived ease of use variable , all indicators also have factor factor loading above 0.5 with the highest indicator value in the " ease to use " indicator of 0.938 and the average respondent's perception value is 4.31. This indicates that each user will use a technology ( e-cash adoption ) if the user feels the ease of use, in which case the respondent uses the e-cash application because the respondent feels ease to use .
All values of perceived of usefulness indicators can also form perceived of usefulness variables because the perceived of usefulness variable indicator has a value greater than 0.5. The " improves banking productivity" indicator has the highest factor loading value of 0.918 with the average respondent's perception value of 4.35. This illustrates that these indicators are dominant in representing perceived of usefulness variables. In addition, this condition reflects that most e-cash users use e-cash because they want to increase productivity in banking transactions.
The factor factor loading of attitude indicator indicates that 3 (three) indicators significantly form attitude variable. The " Wise Idea " indicator is the indicator that has the highest factor loading which is 0.940 with the average perception value of 4.06. Furthermore, for behavioral intention to use the variables , " Continue Using " indicator has the highest factor loading value, that is 0.945, with the average value of respondents is 4.03.
Hypothesis testing is done by using t test ( t-test ) on each lane of influence between variables. In the PLS statistical test each hypothesized relationship is performed using a simulation. Testing with bootsrap is also intended to minimize the problem of research data abnormalities. Test results with bootstrapping from PLS analysis can be seen in Table 7 It has been determined beforehand that the value of t-table with 5 percent significance is equal to 1.96. Primary Data, 2018 All path coefficients in Table 7 have a t-statistic value above 1.96 so it can be expressed to have a positive and significant effect. Thus the whole hypothesis is accepted.

V. CONCLUSIONS
The results of statistical data analysis show that e-trust has a significant effect on perceived of usefulness. This shows that the higher the level of user trust in digital financial services, the greater the benefits felt by users of digital e-cash financial services.
The results of statistical data analysis shown that e-trust has a significant effect on attitude towards using. This suggests that the higher the level of trust given by users to the digital e-cash financial services the better the attitude shown by the users to use the digital financial services.
The results of statistical data analysis shown that e-trust has a significant effect on behavioral intention to use. This indicates that the higher the level of user confidence in digital financial services the higher the intention to use the service.
The results of statistical analysis show that perceived ease of use has significant effect on perceived of usefulness. This shows that the greater the perceived ease of use of digital e-cash financial services, the greater the perception of the benefits perceived by users.
The results of statistical data analysis show that perceived ease of use has significant effect on attitude towards using. This suggests that the easier the use of digital e-cash services (the greater the user's perception of ease of use) the user will show a good attitude toward the digital financial services.
The results of statistical data analysis shown that perceived of usefulness has a significant effect on attitude towards using. This shows that the greater the level of user perceptions of perceived benefits, the better the user's attitude towards digital financial services will be.
The result of statistical data analysis shows that perceived of usefulness has significant effect to behavioral intention to use. This suggests that the greater the level of user perception of the usefulness of the digital e-cash financial services perceived, the higher the intention to use the digital financial services.
The result of statistical data analysis showed that attitude towards using has a significant effect on behavioral intention to use. This suggests that the better the user's attitude toward the use of digital financial services the higher the user's intention to use digital e-cash financial services.

VI. SUGGESTION
Further research is expected to examine other variables associated with the intent of use of information technology, or add dimension of e-trust into other dimensions that significantly affect the process of technology system