T he Effect of Online Shopping Experience on Risk P erception and T rust for Building Online Repurchase Intention

Purpose : The purpose of this study is to determine the role of mediating risk perception and trust on the influence of online shopping experience on the intention to conduct online shopping transactions. Design/Methodology/approach : This type of research is quantitative, using pre-existing measurement scales. Findings: The results of this study indicate that the experience of online shopping has no effect on risk perception, online shopping experience has a positive and significant effect on consumer confidence and purchase intentions, risk perception has no effect on consumer purchase intentions, trust does not affect consumer purchase intentions. Practical/Implications: The next researcher is expected to be able to modify the research model by adding and developing indicators and other variables and giving consideration to online service providers to pay more attention to their services in the field of online shopping transactions.


Research Background:
The internet has facilitated transaction activities in various countries including developing countries. Indonesia is part of a developing country in Asia, which is densely populated and is experiencing rapid e-commerce growth (Chen et al., 2017). Statista report in 2018 shows, internet users in Indonesia numbered 132.7 million occupying the third largest ranking in the Asia Pacific Region. This amount contains a large potential in changing consumer behavior related to the way of shopping from conventional manual transactions to online shopping transactions. The survey data results of the Indonesian Internet Service Providers Association (APJII) in 2017 about internet usage based on age in Indonesia obtained the age of 13-18 years as much as 16.68%, ages 19-34 years as much as 49.52%, ages 35-54 as many as 29 , 55%, and age> 54 years as much as 4.24%. The data shows, the demographic characteristics of Indonesian consumers who use more online transactions are young age. ⋆ Corresponding author.
Besides Indonesia, another country dominated by young segments in conducting online transactions is China. Therefore, China and Indonesia have almost the same demographic profile, the majority of consumers are young people (CNNIC, 2014; Purwanegara & Andika, 2013). These two countries have been considered as potential online markets (Chen et al., 2017). Age characteristics are a significant factor influencing online shopping experience and trust (Wan et al., 2012).
A comparative study of young segments in China and in Indonesia on the reasons for online shopping shows a significant difference in the perceived trust and ease of use of online shopping. However, respondents from both countries were generally satisfied with their online shopping experience (Chen et al., 2017). It is interesting if this study is more focused on a particular province in Indonesia because this country is formed from the diverse cultural characteristics of each province.
The subject of the research is the young segment of Denpasar City in Bali which is one of the provinces in Indonesia whose economy is developing rapidly due to the growth of tourism investment. The choice of subject and location of this study is due to the gap in business phenomena in the form of research limitations on the influence of the experience of online shopping transactions for young people in Bali, especially in the city of Denpasar.

Research Gap:
The results of most of the empirical studies on the effect of online shopping experience on intention to shop online show non-uniformity. Some showed significant ( Study of online shopping behavior of the young segment has been carried out in various countries such as the USA, China and Indonesia, expressed several things namely; the existence of trust, and consideration of the level of their perception of risk. In the USA, the negative experience of transacting young consumers online directly increases their concerns about the privacy of their online information (Yang, 2012). The same thing happened in China. The negative experience of young internet users increases concerns about their online privacy thereby increasing the risk they feel and damaging their trust (Yang, 2013). Risks such as insecurity reduce their intention to shop online (Matic and Vojvodic, 2014). This condition is very different from the empirical results that, subjective perception of age has a positive influence on the formation of risk perception (Le Serre et al., 2017). As people get older or get older, people will consider the risks more.
Two factors namely; beliefs and perceptions about risk can mediate the effect of online trading experience on intentions to transact online that has so far been found to be non-uniform.

Formulation research problem:
The whole set of phenomena and research gap presented provides clear directions on research formulations, namely: Does online shopping experience influence risk perception, does online shopping experience affect customer trust, does online shopping experience influence the intention to conduct online shopping transactions, does risk perception influence the intention to do online shopping transactions and whether trust affects the intention to do online shopping transactions.

Online repurchase intention:
The general definition of purchase intention is a component of consumer cognitive behavior, which describes the specific intention to buy a particular product (Ling et al., 2010). Intention to buy is often referred to as a conative component in attitudes related to the possibility of consumers to buy certain products. Purchase intention in the context of online transactions is described by Pavlou (2003) as a purchase intention that occurs when a customer is willing and interested in engaging in online transactions. Meskaran et al. (2013) states, online purchase intentions are the readiness of customers to buy on the online platform. Chen (2012) identified three components of buying interest namely; interest that is immediate or in the short term, long term, and interest in available products. Past experience of satisfying service to repurchase intentions has been widely carried out (

Perceived Risk:
The perceived risk (perceived risk) of consumers in online shopping is defined as consumer knowledge about the possibility of uncertain negative results resulting from online purchases . Schiffman and Kanuk (2010) define risk as the uncertainty faced by consumers when they cannot predict the impact of their purchasing decisions.
The six components of perceived risk associated with spending in general are as follows: physical, social, product, comfort, financial, and psychological risks (Peter & Tarpey, 1975). In the context of online shopping, there are three components related to the risk perceived by customers, namely; perception of product risk, perception of financial risk and perception of privacy risk (Dai et al., 2014). According to Bhatnagar et al. (2000) show three main types of risk: financial risk, product risk, and information risk (security and privacy). Empirical studies prove, risk components that seriously affect the purchase intention are; financial, product risk (Durmus et al., 2017) and privacy (Drennan et al., 2006).
The results of empirical studies prove, there is a negative effect of perceived risk perceived by consumers on the in- . Based on the results of an empirical study of the effect of risk perception on purchase intentions, the following hypotheses can be arranged: H1: Risk perception has a negative influence on consumers' intention to shop online (online repurchase intention).

Trust:
Trust has various definitions. The Oxford Dictionary has 17 definitions for the word trust, while Webster has 18 (Tan

The Effect of Online Shopping Experience on Risk Perception and Trust for Building
Online Repurchase Intention The online shopping experience has a positive influence on customer trust (Tasin, 2017). Poor online shopping experience will cause a decrease in customer confidence in online transactions (Chris, 2012;. Based on empirical findings about the effect of online shopping experience on customer trust, the following hypotheses can be compiled: H4: The experience of shopping online has a positive effect on trust In different studies online shopping experience has an influence on risk perception (Tong, 2010; Dai et al., 2014). Customers with a strong desire to buy goods online and supported by good prior online experience will help them reduce uncertainty (Shim and Drake, 1990). The hypothesis formulation is as follows: H5: Online shopping experience has a negative influence on consumers' risk perception.

RESEARCH METHODS:
This study examines the relationship between online shopping experience variables, risk perception, trust with online purchase intentions so that the research is associative. Inferential analysis techniques are used to test empirical models and hypotheses proposed in this study.

Variables and measures:
There are four variables involved in this study, namely; the online shopping experience of the indicator was adopted from Lin and Lekhawipat (2014), the risk perception of the indicator was adopted from Dai et al. (2014), the confidence indicator was adopted from Chen (2012), and purchase intention was adopted from Chen (2012). Indicators of each construct have been empirically tested and used in previous empirical studies are shown in Table 2 Table 2. All items as instruments in the study used five (5) points Likert scale ranging from "strongly disagree (score 1)" to "strongly agree (score 5) ".

Construct validity and reliability:
A total of 30 respondents were involved in the validity and reliability test. This test is done by finding the significant correlation and Cronbach Alpha of each dimension and construct. The construct is declared valid if it has a correlation coefficient above 0.30 (Nunnally, 1960), and is said to be reliable if it has a Cronbach Alpha (α) value above 0.60 (Nunally, 1960). The results of the validity and reliability test showed, all indicators involved in this study were declared valid and reliable.

Sampling:
The population is young people aged 17 years to 31 years in Denpasar City, Bali Province in Indonesia. The nature of the population is mobile, therefore, determining the number of samples using non-probability. The sampling technique uses accidental sampling technique. A total of 96 young people were included as samples in this study. Table 1 is the characteristics of the respondents involved as a sample.

DATA ANALYSIS:
This research uses the loading factor criterion 0.6 which is considered sufficient, because it is the initial stage of developing the measurement scale and the number of indicators per variable is not large, ranging from three to seven indicators (Ghozali, 2011). The results of this study indicate, there are several first order loading factors that must be eliminated because they are below the criteria (Y2.3 and Y2.6.) That build trust variables. Then the process is repeated to fulfill convergent validity. Table 2 shows the entire value of the second process loading factor.
Second order analysis shows the path value of each risk dimension is proven to be significant as evidenced by the results of the T test statistic greater than 1.96 so that the three dimensions of risk can be expressed as valid dimensions forming risk perception. Table 3 shows Composite reliability is said to be good if it has a value above 0.70. Composite Reliability (Ghozali, 2011) in this study has a value above 0.70, which means it meets the reliable criteria.
The information that can be obtained in the presentation of Table 4 can be seen that of the three variables having AVE values above 0.5 and the √ AVE value is greater than the correlation coefficient between one variable with other variables, so it can be concluded that the data in this study have discriminant validity good. However, the risk perception variable has a AVE value below 0.5 which means discriminant validity is lacking. The structural model is evaluated by observing the Q2 predictive relevance model that measures how well the observed value generated by the model obtained a Q2 value of 0.6560. This value can be interpreted that the information contained in the data 65.60 percent can be explained by the model, while the remaining 34.39 percent is explained by errors and variables not yet included in the model.

5
RESULT: Table 5 shows the results of the hypothesis test. Hypothesis 1 states that risk perception has no effect on purchase intention (β = -0.136; and T Statistics <1.96) this means that hypothesis 1 is not supported. Hypothesis 2 states that trust does not affect purchase intention (β = 0.098; and T Statistics <1.96), meaning hypothesis 2 is not supported. Hypothesis 3 states that online shopping experience has a positive effect on purchase intention (β = 4.478; and T Statistics <1.96) hypothesis 3 is supported. Hypothesis 4 states that online shopping experience has a positive effect on trust (β = 0.626; and T statistic <1.96) means hypothesis 4 is supported. Hypothesis 5 states that online shopping experience does not affect consumer risk perception (β = -0.183; and T statistic <1.96) hypothesis 5 is not supported.

The Role of Mediation:
The mediation variable testing method used is using  Table 6 shows the results of testing the role of mediation. Risk and trust perceptions do not mediate online shopping experience with the intention to conduct online shopping transactions. It seems that millennials do not consider risk and trust in conducting transactions online. Their main factor for making transactions online is their experience of making transactions online. This condition is proven by trust and risk perception having no role in mediating the online shopping experience of online repurchase intention.

DISCUSSION:
The majority of buyers who use online transactions are in a younger age group because they are more responsive and innovative to technological advancements (Ganguly et al., 2010). Although there is an assumption that the risk of online shopping is higher than offline shopping (Forsythe and Shi, 2003), the results of this study indicate that young people do not consider risks in shopping online.
The results of this study support the general assumption of a variety of literature that, young people do not consider too much risk in decision making. This also occurs in the context of online transactions as indicated by the insignificant role of perceived risk mediation in the influence of online shopping experiences on the intention to conduct online transactions. They only involve direct experience. The better their previous experience in conducting transactions online, the greater the opportunity to intend to do shopping transactions online in the next period. This condition can be the cause of millennials having a behavior that contributes strongly to online shopping (Dhanapal et al., 2015). Millennials in most regions of Indonesia and Bali have economic characteristics that are not yet well established so that their transactions are still nominally limited, so they do not consider risk too much (seeTable 1 Characteristics of Respondents).
Interesting finding. Although their experience in conducting online transactions affects their level of trust in an online shop operator, but to build an intention to conduct online shopping transactions the young generation does not involve trust in a particular online shop organizer. That is, they decided based on direct experience with the next online store without considering the brand of the online store to build their intentions. The online shopping experience has no proven effect on consumer risk perception. The level of experience possessed by consumers in conducting online transactions has no effect on the level of risk perception of consumers in conducting online transactions. This finding shows young consumers do not think about and consider the risks from their experiences. Society in Indonesia, especially in Bali, has a consumptive nature, so whether or not someone's experience in online transactions will not affect the level of perceived risk perceived when conducting online shopping transactions. This finding does not support the results of empirical studies in other countries which state that the more experienced a person is in shopping online, the lower the risk perception    Risk perception has no proven effect on consumer purchase intentions. This means that the perceived risk perceived by consumers does not determine the intention to spend where the results of this study are in line with the study of Santoso et al (2017) who found that risk perception has no effect on buying interest. And these results are different from previous studies which found that risk perception has a negative and significant influence on purchase intentions (Martinayanti et

Managerial and research implications:
The results of this study become a new reference to clarify previous research that, the more experienced people in shopping online will increase the intention to conduct online shopping transactions. Although the character of young consumers does not consider risk, online store operators must be careful, because their experience of online web stores greatly affects their trust in the types of online store operators, therefore the reputation of online stores remains a major factor.

Limitations and future lines of research:
The distribution of questionnaires in this study, using the online application form so that researchers do not directly meet with the response in the distribution of this questionnaire. The sample used is still relatively limited, further research in the future needs to use more sample involvement in longer research time. This study uses an existing measurement scale so that it is less representative of culture in Indonesia, especially Bali, which has specific characteristics.
In addition, researchers can then modify the research model by adding and developing indicators and other variables. Future research, it is strongly recommended to predict the effect of risk on the level of trust that is not measured in this study.