INVESTIGATING CUSTOMER TRUST AND POSITIVITY OF PERCEIVED CUSTOMER SERVICE SKILLS ON THE CONFIDENCE OF ONLINE SHOPPING

With the economy's continuous development, consumers become more mature and rational, so many keen enterprises gradually shifted their focus from the initial product development to "customer service skill" and thus obtained a good market return. Especially for the online shopping activities carried out in the virtual (digital) way, due to the particularity of its channel, it is more important to study the experience of its customers in online consumption. In addition, shopping websites and businesses selling goods through the internet should try their best to use words, pictures, video, and other media, as well as setting up more friendly interfaces. This paper takes the telepresence, pleasure, social presence, and trust in the process of online shopping as the research variables. Through the stepwise regression analysis method, it is found that telepresence can well predict the pleasure of customers in the process of online shopping. However, as the research on customer trust and positivity of perceived customer service skills is scarce, this paper can help to understand the impact of customer satisfaction on services' behavior intention, such as trying to increase the opportunities for communication with consumers and help them during the shopping process, so as to effectively increase consumers' trust in websites and businesses.


INTRODUCTION
With the gradual popularization of Internet technology, the increase in user operation proficiency and the acceleration of the pace of life, the world's online shopping market has developed rapidly (Islam, 2021). The number of online purchases and the quantity of time spent online have increased (Monsuwe et al., 2004). The deepening of people's understanding of economic activities, and trust, which has always appeared as a silent factor in addition to formal incentives or reliable constraints, has gradually attracted extensive attention since the 1980s (Rodrik, 2005). Individual trust in other people's values is measured by their assumed honesty, virtue, and competence (Eneizan et al., 2020). An increasingly strong understanding is being formed: trust, as the primary social capital, plays an essential role in the construction and operation of the social economy (Ye et al., 2005). In this context, the trust issues that are mainly concerned with philosophy, ethics, sociology, and social psychology have returned to the theoretical vision of economics and management (Strub & Priest, 1976). Trust can go beyond the constraints of laws or contracts, reduce customer-perceived risks more effectively than any regulations, and help customers improve their emotions (Guo et al., 2021).
Compared with continuous trust, although initial trust is not a stable psychological state, it shows the courage of both parties to complete online transactions and has an important influence on the formation of continuous trust (Tao et al., 2021). The initial trust establishment stage is also a key stage for potential customers to understand online sellers and decide whether to trade with them (Lin et al., 2014). According to a survey conducted by iResearch and Consultancy Company in China, during the e-commerce shopping festival in 2015, the payment conversion rate from the product page to the payment page is divided into 6%, 11% and 14% on Taobao, JD and T-mall, respectively, which means that even if the promotion efforts are increased, the vast majority of customers still abandon the purchase after browsing the store. It can be seen that how establishing solid customer trust in the initial stage and transforming shop viewers into actual purchasers has become one of the most concerning issues for online sellers especially subjective online word-of-mouth, to judge product quality (Talwar et al., 2021).
Websites are critical to a company's success because they serve as a communication route between the company and its customers (Chen et al., 2016). Although many shopping platforms and shopping websites are trying to improve the security of online transactions through technical means and have achieved certain results, the problem of website trust has not been really solved, which has always been one of the most important contents in the research of e-commerce by scholars. Because it is of real significance to talk about customer experience and satisfaction only after truly solving the problem of trust. Therefore, solving the problem of network trust is the key to solving the problem of follow-up network consumption behavior.
Although scholars have different summaries of these influencing factors, the factors that are more consistent and recognized can be directly controlled by shopping websites and formulate corresponding procedures to implement them (Armstrong & Overton, 1977). The main three factors that are improved are the convenience of use, transaction security, and service sincerity. Therefore, this article will focus on the impact of these three elements on customer trust in online shopping, compare their relative influence, and examine the moderating effect of shopping websites' commitment violations on the customer trust mechanism (Chaouali & el Hedhli, 2019).

LITERATURE REVIEW
The research on trust in the website environment, especially the trust for online consumers, began in the middle and late 1990s. It can be roughly divided into three stages: early, middle, and late. Many other scholars have discussed the factors affecting consumers' trust building from the aspects of consumers' characteristics, website characteristics, commodity characteristics, service quality, evaluation system, and external environment (Jiang & Lau, 2021). At the same time, some scholars study the impact of trust on consumption behavior from another perspective. For example, Song et al. (2021) found that trust has a significant impact on online shopping preferences through data analysis. Chen and Yang (2021) found that consumers' trust experience can play a positive role in optimizing the customer experience of the website through the analytic hierarchy process.
Regarding the initial trust of online customers, McKnight et al. (2002a) conducted research for the first time and constructed a conceptual model of initial trust. The research found that the customer's trust tendency, trust belief, and trust based on system and cognition will lead to his initial trust in the seller. Customer trust in the platform is critical since it encourages customers to use it (Chaouali & el Hedhli, 2019). McKnight et al. (2004) and McKnight et al. (2002b) constructed an initial trust establishment trust-building model (TBM) in their follow-up research. The empirical study found that the seller's reputation, website quality, and structural guarantee have a positive effect on the customer's initial trust and influence. Wakefield et al. (2004) focused on the impact of website quality and structural guarantees on the initial trust of customers based on the TBM model and found cognitive cues that represent website quality, such as communication, opportunism, brand equity, website attraction, etc. Significantly affect the initial trust of customers, and the structural guarantee of the website also helps to increase the initial trust of customers but does not affect customers' willingness to purchase.
Subsequently, scholars successively paid attention to the influencing factors of customers' initial trust and constructed different models of initial trust influencing factors. In a nutshell, the influencing factors include three aspects: website/shop factors, customer personal factors, and third-party factors.
From the perspective of information theory, Shao et al. (2021) first put forward the concept of "clue", defined it as the information used to evaluate and compare different products or services, and pointed out that products are composed of a series of clues (such as price, brand name, packaging, color, etc.), and customers' cognition of product quality is a collection of perception of various clues of products. Steenkamp (1990) believes that cues are conscious or unconscious, verbal, or nonverbal information that can be observed by individuals and will affect their decision-making process. Generally speaking, the clue is the information released by the encoder and accepted by the decoder as the product evaluation standard. Brehmer (1972) divided product leads into two dimensions: predictive value and confidence value and believed that customers often rely on these two dimensions to judge the product quality. These two dimensions are used to help customers increase the correctness of purchasing decisions and enhance purchasing confidence. Predictive value refers to the degree to which customers are convinced that a certain clue is related to product quality and regard it as a product quality indicator. Therefore, Chang and Chen's (2009) confidence value refers to the degree of confidence that customers can use a certain clue to correctly evaluate product quality. Nakayama (2017) studied the purchase situation of experiential/hedonic products and found that customers rely more on product external clues, especially subjective online wordof-mouth, to judge the product quality. At this time, product external clues have higher predictive value and confidence value for customers.
The stimulus-organism-response model proposed by Mehrabian and Russell on the basis of environmental psychology, that is, the S-O-R model, was first used to analyze the retail environment (Lee et al., 2011). Nowadays, it is increasingly used by scholars to analyze the environment of online shopping. Based on the connotation expression of the model, when consumers have different experiences due to the influence of a certain stimulus (S) from the website, the body (O) will produce different perceptions, attitudes, and emotions, so as to form a new cognitive structure, respond (R) under the common action of stimulus and new cognitive structure, and finally form an overall attitude towards the website (positive/negative, trust/distrust).
Customers prefer to buy online since it is more convenient and saves time (Huseynov et al., 2016). Customer online shopping experience can be summarized as a kind of customer experience and feeling in the process of online consumption interaction (Verhulst et al., 2020), a multi-dimensional and personal psychological state, including both emotional experience and cognitive experience (Punyatoya, 2019). Website is a psychological state of an individual's subjective perception of the environment, and the customers' loyalty to online shopping will increase through image quality of goods such as 3D view and image watermarking (Nawaz et al., 2020). It is an immersive subjective experience obtained by the individual through automatic or controlled psychological processing.
It can be seen that the research on the issue of online consumer trust has always been the focus of attention of scholars, and the research methods have gradually shifted from qualitative research to quantitative research, and the objects of research have also changed from focusing on corporate management and website construction.

Trust Influence Mechanism
Clear product classification catalog, efficient and fast user interface, and shopping process have differentiated commodity recommendations according to customers' purchase preferencethese convenient services of shopping website save customers' time cost of search and browsing (Fuentes & Svingstedt, 2017), so that customers can obtain a large amount of product and price information in a short time, and then conduct comparison and screening to select the most satisfactory products (Shehzad et al. 2021). Online retail sales hit $US2.8 trillion in 2018 and are expected to reach $US4.8 trillion by 2021(Coppola et al., 2022. Moreover, these convenient services also prompt customers to recognize the professionalism and comprehensiveness of the shopping website and make customers believe that the shopping website has invested in product storage, process design and customer tracking services, and they hope to maintain good cooperation with customers. Previous research has indicated that older persons are more inclined to use and adopt online technologies if they see their utility and benefits, such as with health care (Hoque & Sorwar, 2017). The relationship thus enables these investments to obtain continuous returns and will not take vicious fraud to destroy the cooperative relationship between the two parties and only obtain short-term petty profits. To sum up, the convenience of using shopping websites can make customers have confidence in the integrity and ability of shopping websites and improve customers' trust in shopping websites. From this, we propose: Hypothesis 1: The ease of use of shopping websites can increase customers' trust in shopping websites.
Website privacy/security features, as well as shared values, are important aspects in online shopping trust. Due to the particularity of the online shopping process, customers have to pay in advance to the seller or a third-party regulatory agency; upload their personal information and account information to the website; cannot see the actual product and can only understand the product through text descriptions or pictures. In this way, once the seller does not deliver the goods, the description of the goods is inconsistent with the physical object, and the customer's personal information is stolen or tampered with, it will not only cause the customer to suffer economic losses, but also bring a lot of inconvenience and trouble to the customer's daily life. Confidence is one of the most crucial factors influencing the success or failure of online shops (Prasad & Aryasri, 2009). Therefore, the security of online transactions has become one of the key factors restricting the development of online shopping in my country (Woltjer et al., 2021). If the shopping website can formulate processes, norms and dispute resolution mechanisms to protect customers' rights and interests, it will bring a psychological guarantee to customers so that customers know that they also have the right to control the payment and the whole transaction before receiving the product, ensuring that the product quality is consistent with the commitment and confirming it. Shodiq et al. (2018) emphasized in their research that clients should be convinced by ensuring their security and an immediate payment transaction process. Once there is a problem, they can recover their advance payment through corresponding procedures to protect their own interests; know why you want to provide personal information, how this personal information will be stored and contacted by whom, and how you can stop it in case of theft and tampering. This will significantly eliminate customers' worries in online shopping and allow them to make transactions with confidence.
From this, we propose:

Hypothesis 2: The transaction security of shopping websites can improve customers' trust in shopping websites.
Similar to the traditional sales model, service sincerity has also become an effective means for shopping websites to compete with each other and cultivate customer loyalty. Due to the particularity of online transactions, customers always keep a certain distance from shopping websites, and there is little direct communication. Online shopping satisfaction refers to the user's reaction to the output of an information system, which may be measured by the dimensions of repeat purchases and repeat visits (Husain, 2019). Sincere service shortens the psychological distance between customers and shopping websites and makes customers feel that shopping websites actually put themselves in the position of a service provider. Before making a purchase, customers always look for competitive pricing and compare them to other online stores (Hamza & Saidalavi, 2014). In addition to pursuing economic interests, they also care about customers' personal development and focus on cultivating long-term friendships with customers. By providing high-quality services, most online selling platforms develop customer confidence and establish long-term partnerships (Shafiee & Bazargan, 2018). In this way, once customers perceive and confirm the sincerity of the shopping website, they will regard the shopping website as a helper and friend in life, believe in the shopping website, and believe that the recommendation of goods and personality sold by the shopping website is out of the good intention of helping themselves improve their life, rather than being driven by economic interests. Therefore, we propose:

Customer Buying Behavior
Despite the worldwide increase in internet usage and online shopping, customers in some nations remain hesitant to shop online due to a lack of confidence (Stouthuysen et al., 2018).
The unique uncertainty and risk of online shopping bring a great concern to customers. Trust can eliminate these concerns (Eneizan et al., 2020). Jeon et al. (2017) studied the function of online trust and utilitarian value in mediating the relationship between perceived website engagement and the online travel community. When customers trust a shopping website, on the one hand, customers will have confidence in the performance ability and commodity quality of the shopping website and think that the commodities they buy will not fail and are worth their money; On the other hand, customers will also recognize the business reputation and transaction sincerity of the shopping website and believe that the website will not take speculative or immoral acts to damage their own interests. These two points together help customers resolve their concerns about the potential risks of online shopping, enhance customers' confidence in online shopping, and effectively promote customers' purchase behavior. Therefore, we propose:

The Shopping Website's Commitment
Compared to the traditional sales model, the initiative and controllability of customers in online shopping is weaker, and the confidence of customers in online shopping comes from the good protection of customer rights and interests of shopping websites. Commitment is an essential component of a long-term relationship's success (Dwyer et al., 1987). Therefore, once the shopping website violates its commitment to customers, it will have a huge negative effect on the customer's trust in the shopping website and then the implementation of the purchase behavior. According to one study, dedication is an effective and long-lasting predictor of behavior (Lokhorst et al., 2011).
If the shopping website violates its commitment to customers, even if the shopping website continues to provide customers with convenient browsing and retrieval interfaces, safe transaction processes and sincere services, it will make customers doubt the authenticity and authenticity of these services and processes. Most probably, it is believed that these processes and promises are difficult to achieve. Once transaction disputes arise, it is difficult to protect their own rights. The website may make these promises to set traps to induce customers to purchase and to obtain benefits from them, which will make customers feel better about the shopping website. The trust is greatly reduced. From this, we believe that:

Hypothesis 5: Commitment violation behavior of shopping websites has a negative moderating effect on the positive relationship between customer trust influencing factors and customer trust.
Based on the above discussion, we proposed the corresponding conceptual model and hypothesis, as shown in Figure 1 below.

RESEARCH METHODOLOGY
In this research, we collected data in the form of a questionnaire. The questionnaire is divided into two parts (a total of 28 items): the first part is the variable measurement, including telepresence four questions; social telepresence four questions; six items of pleasure; sixteen items of trust; all variables are measured using the 7-Likert scale. The second part is background information, which collects the surveyed data, including gender, age, education level and monthly income. The Statistical Software SPSS (Statistical Package for the Social Sciences) was used to analyze the data and check the reliability of the questionnaire.

Analysis of the Reliability and Validity of the Questionnaire
A total of 390 questionnaires were sent out through both online channels and face-to-face surveys, and 366 valid questionnaires were actually received. Through the test, the value of each variable is greater than 0.6, so the questionnaire is reliable for testing the relationship between the independent variable and the dependent variable. At the same time, the principal component analysis method is used to intercept the data with the characteristic root value 1 as the standard, and the orthogonal rotation to maximize the variance (Varimax) is used for exploratory factor analysis. Each question item is clearly loaded on the main factor, and they are all greater than the acceptable level of 0.5, indicating that they have good structural validity.
Among the 28 items in the questionnaire, the highest average value is 4.5820. The corresponding question is "the main online businesses are very clear about the market law".
The highest standard deviation is 1.43464, and the corresponding question is "usually I hope I can understand how online sellers can consider how their own behavior affects me".

Correlation Analysis
The results of the correlation analysis between the variables can be seen from Table 1. There is a positive correlation between telepresence, social telepresence, and pleasure. There is also a positive correlation between pleasure and trust.

Factors Affecting Online Shopping Pleasure and Trust
The collinearity problem of the data is analyzed by variance test. The expansion factor (VIF) is between 1.0 to 1.431. The collinearity problem only expands the variance of the regression coefficient of the independent variable by 1 to 1.4 times, indicating that there is no serious collinearity problem.
First, take pleasure as the outcome variable, and use telepresence and social telepresence as predictors to perform a stepwise regression. The regression results are shown in Tables 2 and   3. It can be seen from Table 2 that social presence has not entered the regression equation, and physical presence can significantly predict the pleasure of online shopping (p<0.05).  In the second step, the degree of trust is used as the outcome variable, and telepresence, social presence, and pleasure are used as predictors to perform a stepwise regression. The regression results are shown in Tables 4 and 5. It can be seen from Table 4 that physical presence, pleasure, and social presence can significantly predict trust in online shopping (p<0.05). It can be seen from Table 5 that pleasure has the most significant influence coefficient on online shopping trust (0.153), followed by the social presence (0.124) and physical presence (0.117).