Effect Of Ease of Use and Trust on Purchase Decisions Through Online Shopping Sites

This study aims to determine: (1) the effect of ease of use on purchasing decisions through online shopping sites (2) trust in purchasing decisions through online shopping sites (3) the effect of ease of use and trust on purchasing decisions through online shopping sites in the people of Tangerang City. The samples taken were 100 respondents. The data collection technique uses a questionnaire that has been tested for validity and reliability. The data analysis technique used is multiple regression. The results of this study indicate that: (1) there is a positive influence of ease of use on purchasing decisions through online shopping sites, as evidenced by the t valuecount of 9,170 a significance value of 0.000 <0.05; and a regression coefficient of 0.687, (2) there is a positive influence of trust on purchasing decisions through online shopping sites, as evidenced by the t valuecount of 3,539 a significance value of 0.001 <0.05; and a regression coefficient of 0.267, (3) there is


I. INTRODUCTION
With the development and technological innovations in the form of sophisticated electronic media such as smart phones and computers, there is a potential for the creation of electronic commerce or better known as online shopping. For now, in Indonesia itself, there are many online shopping sites.
Online shoppingitself is the distribution, purchase, sale, marketing of goods and services through electronic systems such as the internet or television, www, or other computer networks (wikipedia.com). With online shopping, it will certainly make it easier for people to get a product or service because it will save time and energy.
In many cases, there are various determinants of the success of online sales, including online sales service management companies or online sites that must be able to maintain their existence in the field of online sales, by not only focusing on product sales but also paying attention to reliable and professional company management, accuracy delivery, excellent service, and ease of use starting from the initial opening of the site, registration to the transaction.
Technological advances that are increasingly developing rapidly have an impact on all aspects of human life, including the development of the internet which is currently growing and giving rise to various modern innovations. In this particular case, it is in the aspect of electronic commerce or better known as online shopping. In online shopping, of course, there are various factors that support and support transactions. Because as is known, in online shopping, of course, there is no direct meeting between the seller and the buyer. Therefore, researchers are interested in conducting research on "The Influence of Trust and Ease of Use on Purchase Decisions Through Online Shopping Sites with case studies on people in Tangerang City".

II. RELATED WORKS/LITERATURE REVIEW (OPTIONAL) EASE OF USE
Basically, this convenience factor is related to how operational transactions are done online. Perception of convenience in online shopping itself can be interpreted as the extent to which a person believes that using a technology will be free of effort. Indicators of the ease of use variable include: 1. Easy to learn (easy to learn): It is easy to learn a site with its features, especially for new users. 2. Controllable: Every administrative data and order data can be changed by the user (customer) according to their wishes and needs with an agreement that has been agreed upon by the buyer and the seller (online shopping site) beforehand. 3. The quality of information that is clear and understandable: It is the customer's perception of the quality of information about products or services provided by a website 4. Flexible (flexyble): It is the desire to develop and adopt new technologies that will be added value and then embed it into the site. 5. Time efficiency: Time spent in carrying out an activity, especially activities in using the site. 6. Easy to become skilled (easy to become skillful): Proficient according to the Big Indonesian Dictionary is defined as being highly trained in doing something. In other words easy to become proficient meant one's ability in running a site. 7. Easy to operate as desired: Ease of use according to what the user or users want. 8. Easy to use: Ease of use that does not require much effort in running a system. 9. Ease of payment transaction system: It is a system that includes a set of rules, institutions, and mechanisms used to carry out the transfer of funds to fulfill an obligation arising from an economic activity. 10. Ease of interaction with sellers: There are features that facilitate communication between buyers and sellers, for example to buyers when there are complaints about products.

TRUST
Trust is the first factor that becomes a consideration in buying a product. This is because before the product is purchased by the customer, the producer must be able to create customer trust in the product being offered, the goal is to attract attention and generate customer interest and confidence in the product being marketed. If consumers trust the online site provided by the company, then it allows them to increase their desire to make purchases online. In addition, consumer confidence in deciding to purchase is also found in the popularity of the online site. The more popular an online shopping site is, the more consumer confidence in the online site will increase. Until the next stage is consumer confidence in the security and privacy of consumers towards online shopping sites and sales partners who sell on these online sites.

BUYING DECISION
Decision is a problem solving activity which will result from solving the problem in the form of actions or opinions based on several alternative solutions (Yanti et al., 2022). According to Kotler & Keller (2012: 188), the purchase decision is one of the stages of evaluation carried out by a consumer on the product to be purchased. Purchase decision is an action taken by consumers with certain backgrounds in determining the purchase of a product based on existing alternatives. Based on the data above, it can be seen that there are two gender characteristics, namely male and female. Where the number of female respondents dominates, as many as 52 people with a percentage of 52%, while respondents with male sex are 48 respondents with a percentage of 48%. Based on the data above, the characteristics of the respondents are classified into four groups. As for the 100 respondents, the age group under 20 years is 13 people with a percentage of 13%, the age group with an age range of 21 to 30 years is 61 people with a percentage of 61%, the age group with an age range of 31 to 40 years is 19 people with a percentage of 19%, and the age group above 40 years there are 7 people with a percentage of 7%. Based on data on the length of internet use per day, there are three groups of characteristics. Of the 100 respondents, the longest duration of internet use per day was in the group of more than 3 hours of use, namely 47 respondents (47%). While the group with the lowest number was in the group with duration of use under 2 hours, namely 18 respondents (18%). From table case processing summary it can be seen that the overall data derived from the results of the questionnaire regarding the convenience with the number of respondents as many as 100 respondents, none of them were excluded from the analysis (excluded). The results of the reliability test indicate that the convenience variable has Crombach's alpha above 0.60 so it can be said that the items measuring the convenience variable from the questionnaire are reliable.  (0,1966) with the output correlated item total correlation. 3. If you compare rtable with rcount in the correct item total correlation column, the value of rcount regarding confidence is greater than the value of rtable, meaning that all statements are declared valid. From the case processing summary table, it can be seen that the overall data derived from the results of the questionnaire regarding trust with the number of respondents as many as 100 respondents, none were excluded from the analysis (excluded). The results of the reliability test indicate that the use of trust variable has Crombach's alpha above 0.60 so it can be said that the items measuring the confidence variable of the questionnaire are reliable.  (0,1966) with the output correlated item total correlation. 3. If you compare rtable with rcount in the correct item total correlation column, the value of rcount regarding confidence is greater than the value of rtable, meaning that all statements are declared valid. From the case processing summary table, it can be seen that the overall data derived from the results of the questionnaire regarding convenience with the number of respondents as many as 100 respondents, none of them were excluded from the analysis (excluded). The results of the reliability test indicate that the purchasing decision variable has Crombach's alpha above 0.60 so it can be said that the items measuring the convenience variable from the questionnaire are reliable.  (0,1966) with the output correlated item total correlation. 3. If comparing rtable with rcount in the correct item total correlation column, the value of rcount regarding purchasing decisions is greater than the value of rtable, meaning that all statements are declared valid.  From the data above, it can be seen that: 1. The magnitude of the relationship between the ease of use variable (X1) is indicated by the correlation coefficient value of 0.834, and the magnitude of the relationship between the trust variable (X2) is indicated by the correlation coefficient value of 0.705, so it can be concluded that the ease of use (X1) and trust (X2) variables have a strong relationship. positive. 2. The relationship between the occurrence of variables of ease of use (X1) and trust (X2) which is significant or not with purchasing decisions (Y) can be seen from the probability number which is <0.05, so there is a significant relationship between the three variables. A significant value of 0.000 indicates that the results of the correlation of the three variables are significant, meaning that H0 rejected, and Ha received. From the table above, it can be seen that trust and ease of use are input variables because the probability of trust and ease of use meets the probability criteria, where the value of the probability is below 0.05. If the ease of use variable increases while the ease of use is considered constant, it will cause an increase in purchasing decisions 3. X2 coefficient = 0.267 If the trust variable increases while trust is considered constant, it will cause an increase in purchasing decisions of 0.267.

Hypothesis test F Uji test
The results of the f test were obtained using the ANOVA (Analysis of Variance) table with the aim of testing whether there was an effect of X1 (ease of use), X2 (trust) on Y (purchase decision).

Hypothesis Formulation H0 : i = 0
This means that the independent variables, namely ease of use and trust, do not simultaneously have a significant effect on purchasing decisions through online shopping sites. H0 : i 0 This means that the independent variables of ease of use and trust have a significant simultaneous effect on purchasing decisions through online shopping sites. 2. Determine the magnitude of significance (α) = 5% or 95% confidence level and determine the degree of freedom (df = V1, V2) where V1 = (k-1) and V2 = (nk) to determine Ftable. So the value of Ftable with a significance level of 5% or 0.05 is 3.09 (df1 = 3-1 = 2; df2 = 100-3 = 97). From the ANOVA results obtained Fcount for model 1 is 131.300 with a significance level of 0.000 which is 0.000 <0.05 and Fcount > Ftable or with data 131,000 > 3.09 then H0 is rejected and Ha is accepted, meaning that the ease of use and trust variables have a significant effect simultaneously on the purchasing decision variable (Y).

Coefficient of Determination
Source: Results of SPSS 21.0 From the table above it can be seen that: 1. Column R for model 1 shows a correlation coefficient of 0.855. This means that the effect of ease of use and trust in use on purchasing decisions is quite strong and positive. 2. The R Square column for model 1 shows an R Square number of 0.730. R Square is also called the coefficient of determination. This means that the effect of ease of use and trust in use on purchasing decisions is 0.730 or equal to 73%, while the rest is influenced by other factors. R Square ranges from 0 to 1, the larger the R Square number, the stronger the relationship between the effect of ease of use and trust on purchasing decisions or vice versa. 3. The Adjusted R Square column shows the coefficient of determination, which is 0.725. 4. Std column. Error of the Estimate for model 1 is 2,343 5. In Table 4.50 (Descriptive Statistics) above, there is a standard deviation of the purchasing decision value of 4.669 which is much larger than that of Std. Error of the Estimate 2.2343. Therefore, this regression model is good to use.

T Uji test
Statistical t-test was used to test the effect of the independent variables, namely ease of use (X1) and trust (X2) on the dependent variable, namely purchasing decisions (Y) by assuming that other variables were considered constant and appropriate. Hypothesis formulation H0 : 1 : 2 = 0 it means that the independent variable ease of use (X1), or trust (X2) has no partial effect on purchasing decisions (Y). H0 : 1 : 2 0 meaning that the independent variable ease of use (X1), or trust (X2) has a partial effect on purchasing decisions (Y).

V. CONCLUSIONS
Based on the results of research and discussions that have been carried out on the people of Tangerang City regarding "The Effect of Ease of Use and Trust on Purchase Decisions Through Online Shopping Sites" the conclusions are as follows.
1. Based on the answers to the questionnaire, raw data was obtained which was then processed using the SPSS version 21.00 application so as to produce the regression equation Y = 2.357 + 0.687 X1 + 0.267 X2. Where based on this equation means that every time there is an increase or decrease in trust by one point, the purchase decision through online shopping sites will increase or decrease by 0.687. Meanwhile, if the ease of use increases or decreases by one point, the purchase decision through online shopping sites will increase or decrease by 0.267. 2. Ease of use has a positive effect on purchasing decisions. This is evidenced by a regression coefficient of 0.687 and a significance value of tcount of 0.000 (sig <0.05). t valuecount obtained is greater than ttable ie 9.170 > 1.984. 3. Trust has a positive effect on purchasing decisions. This is evidenced by a regression coefficient of 0.267 and a significance value of tcount of 0.001 (sig <0.05). t valuecount obtained is greater than ttable ie 3.539 > 1.984. 4. Ease of use and trust simultaneously affect purchasing decisions. This is evidenced by the significance value of f of 0.000 (<sig 0.05). f Nilai valuecount obtained is greater than ftable ie 131,000 > 3.09. 5. Based on the magnitude of the simultaneous R2 array in this study, it was 0.730 (73%). This shows that purchasing decisions are strongly influenced by ease of use and trust with a percentage of 73%, and the value of 27% is influenced by other factors not included in this study.
Based on the conclusions obtained from this study, suggestions are proposed as a complement to the research results that can be given as follows: 1. Further researchers can develop this research by conducting research on other factors that can influence purchasing decisions with the aim that the results of the information obtained can be more complete and varied.