Para-Social Interaction and Trust in Live-Streaming Sellers

Live streaming is one of the modern methods that allows sellers to create, transmit, or broadcast some content on the internet in real-time, and it has been used by many small individual merchants. Understanding how live streaming contributes to online consumption is becoming increasingly important in social commerce as the live-streaming industry has grown more and more popular. However, the number of studies on live streaming is still quite limited in Vietnam. Therefore, this research will look at the mechanism that enables live streaming to boost customer trust in streamers. Using PLS-SEM on a sample of 360 respondents who viewed selling live streams on social network sites in Vietnam, we discovered that other members' endorsement, value similarity, hedonic value, and utilitarian value contribute to good para-social interaction. Next, utilitarian and hedonic values, streamer product expertise, and para-social interaction all positively affect trust in the streamers. The findings could help live-streaming sellers better understand their social interactions with viewers, resulting in increased customer trust.

However, the majority of products advertised through live-streaming channels in Southeast Asia are small brands from individual sellers [7].Therefore, we only analyze small sellers or individual sellers in this study since they outnumber large enterprises but are understudied in existing s-commerce research.In reality, there may be many potential risks associated with purchasing from individual sellers via live streaming.Even some well-known live streamers have been accused of deceiving themselves while marketing products.More especially, trust in commercial live streaming differs from trust in other online buying contexts due to specific characteristics such as presentation, quick interaction, and high.In particular, previous research examining trust in online shopping activities has only looked at functional advantages such as quality, confidentiality, image, and usefulness without exploring hedonic incentives as predictors of trust [8,9].Hence, it would also be worthwhile to investigate how live-streaming interactions, utilitarian value, and hedonistic value aid in establishing consumer trust.With Vietnam's rapid development of mass retailing putting livestreaming at the vanguard of retail innovation, we do this research to add knowledge to some future research related to social commerce as well as to provide lessons for other developing countries.
The rest of this paper is structured as follows: The next section will go through the theoretical foundation, research hypotheses, and research model.The following part will describe the research methodology.The findings, as well as their theoretical and managerial implications, limitations, and future research, are reviewed in the last part.

2-1-Live Streaming Business
The live streaming business is popular social commerce in which real-time engagement via live broadcasting facilitates online buying and selling [4].Live streamers can show products and provide viewers with special deals.When customers are watching live streams, they may send messages directly by typing texts to sellers and other viewers.Besides, customers can purchase goods while the live streams are being broadcast.They may buy whatever they want without interfering with the live stream.Live streaming commerce is quickly becoming a fascinating and dynamic online sales channel that overcomes the difficulties of traditional selling online due to its entertaining and real-time nature [1].

2-2-Para-social Interaction (PSI)
The term PSI was used to describe an individual's "illusion of a face-to-face relationship with a media figure" [10].At the time, PSI was viewed as a one-way relationship established between media users and media actors via advertisements, interviews, or television shows.For example, customers tend to consume more when they watch a TV shopping program that has the same host [11].The watchers, who used to connect or had a particular liking for a TV host, are more inclined to make unplanned purchases [12].Para-social interaction, in contrast to interpersonal relationships or involvement, involves a considerably weak link.
Nowadays, the usage of social media for online conversation strengthens users' PSI with other community members [13].PSI has been an important factor in some studies related to customer behavior in online commerce [14][15][16], which is another type of two-way conversation.People can have more engaging and two-sided conversations on social media than they do on traditional mass media platforms [17].

2-3-The Enhancing Impact of Value Similarity
Value similarity refers to how similar individuals engage with each other in terms of beliefs, education, social position, and so on [18].Similarity refers to the status in which people tend to greatly like others who have characteristics that are similar to their own.This concept has received a great deal of attention in the fields of psychology and marketing.According to research, when people connect with people who are similar to them, their communication becomes smoother.If people recognize others who have the same interests, objectives, behaviors, or lifestyles, it is value similarity [19].In this research, "value similarity" is identified as the degree to which social media users have the same interests, beliefs, and preferences as others.
The link between value similarity and para-social interaction has been investigated.Sociologists have shown that in social relationships, people tend to search for people who have similar views or interests to connect and communicate [20,21].Similarity can influence subconscious actions [22].Similarity is very important in the development of social interaction in the sphere of media and communication [23].Online consumers may easily find other members who share their product or service interests, lifestyles, or purchasing experiences.As a result, the study shows that customers' parasocial interaction is favorably influenced by similarities.

2-4-The Role of Other Viewers' Endorsement
Via live streams, a viewer can see the reviews posted by other viewers, which are regarded as other viewers' endorsements.The perception of other users can seduce para-social interactions with live streamers.Personal endorsement refers to viewers' agreement with information transmitted by live streamers [24].Positive comments, likes, recommendations, and follows frequently show how members feel about live streams [25].Word of mouth or social influence is a powerful method to attract consumers to experience or engage.Positive feedback from satisfied customers might help reinforce potential customers' perceptions.Good customer impressions can improve viewers' social engagement with streamers [26].As a result, we offer the following hypothesis.

H2:
The other viewers' endorsement positively influences para-social interaction.

2-5-Utilitarian Value
Utilitarian value reflects how useful or beneficial the live streams are for the consumer.This value can be achieved via live streams when consumers find the product they need, save money, time, and effort, and conveniently access, search, and transact.Customers can also read reviews or engage with one another via live streaming, which satisfies their information needs and improves the outcomes of future purchase decisions.The utilitarian value of live streaming may be seen in its authenticity, reactivity, and visualization.
Customers are concerned with the reputation and authenticity of online sellers [27].This is especially true when the sellers are individuals without a physical store.The term "authenticity" relates to the genuineness, actuality, and originality of anything [28].Live streaming allows customers to view the faces and expressions of merchants, points of sale, and items without having to wait for them to be pre-edited.Live video is said to be more realistic than television advertisements because it is a more realistic representation of what the seller is offering [29].
Second, when purchasing online, customers cannot touch, check, or try items [30].As a result, some websites have incorporated interactive visual technologies to assist online shoppers in experiencing products better [30,31].In the event of live streaming, many sellers have tried the product while live streaming to display the quality as well as instructions for use, making it easier for customers to visualize, assess, and make decisions.
Third, in traditional electronic commerce, clients and sellers are separated in time and space.As a result, responsivenessthe willingness to reply quickly to customer requests and concerns [32] -is critical for increasing service quality and meeting consumer needs.Live streaming has been shown to improve customers' online experiences and behavioral intentions [33].Therefore, the following hypothesis was formulated:

2-6-Hedonic Value
Hedonic value is frequently tied to the level of joy felt by shoppers [34].For example, watching a seller demonstrate or utilize a product, which can be entertaining and thrilling, similar to watching a TV show, can help alleviate boredom or reduce stress.Some online sellers even offer discounts to viewers.Receiving such a deal excites customers [35].Many sellers also design games or activities in which consumers may interact with the live streamer and win rewards.Shoppers may be thrilled to see if they are the lucky recipients of this present.Social media also has a lot of features that encourage a joyful, engaging customer experience, such as on-screen visuals and special effects.The widespread use of mixed media for entertainment and communication has resulted in greater para-social relationships with online streamers.Hedonistic value might be viewed as a latent antecedent of social acts on Facebook [36].Hence, the following hypothesis was proposed:

2-7-Trust in Sellers
The notion that sellers would behave consistently, provide high-quality products, and not cheat customers is referred to as trust in sellers [37].Because of the nature of e-commerce, there is information asymmetry and transaction risk, specifically confusion about the identity of sellers and fears about their opportunism, as well as doubt about product quality [38].Online trust is mostly determined by consumers' perceptions of online sellers giving accurate information and deliver as expected [39].
Utilitarian values such as reputation, perceived security, service quality, and usefulness tend to benefit online commerce trust [8].Through live streaming, utilitarian value persuades consumers that the information presented is real and that they can trust the seller's recommendations and promises.Live streaming reduces identity and product uncertainty, which enhances utilitarian value.Thus, customers should have more confidence in the seller.

H5: The utilitarian value has a positive effect on customers' trust in streamers.
Trust in online sellers is based on the emotional link between customers and sellers [7].Participating in seller activities via live streaming can bring hedonistic value by making the purchasing experience more joyful and engaging.Creating such good feelings might help to establish close connections with sellers and their products.Sellers' hedonistic attempts favorably increase consumers' trust in sellers [40].Live streaming allows sellers to display their items in their unique fashion, which may boost consumer emotions and lead to buyers trusting the sellers and the products.

H6: The hedonic value has a positive effect on customers' trust in streamers.
Via live streaming, customers and sellers interact as if they were friends, which will enhance shoppers' trust in live streamers.It will assist customers in inquiring about, following, and commenting on the content published by other viewers; this capacity encourages them to develop para-social interactions.Also, streamers can instantaneously communicate with audiences, exchanging opinions at the same time, generating good emotions and warm feelings, and reducing customers' skepticism [1].
H7: Para-social interaction has a positive relationship with customers' trust in streamers.
Live streamers may promote certain goods or brands to viewers by talking and sharing information [41].They act as information providers while live streaming.A strong understanding of the product's characteristics may be regarded as a trustworthy source for promoting that product."Product knowledge, competence in utilizing the product, responsiveness to users, and professionalism in their posts" are required of live streamers [42].Product knowledge and streamer expertise are essential for establishing confidence [39].Thus, the following hypotheses are proposed: H8: Product knowledge of streamers has a positive relationship with customers' trust in streamers.

3-Methodology
Proposed research model and the flowchart of the research methodology that was used to achieve the study's aims is shown in Figures 1 and 2.

3-1-Measurement and Data
The theoretical basis is critical for developing knowledge of potential factors related to live streaming.We utilized 7point Likert scales (from "strongly disagree" to "strongly agree") for the measurement items of the constructs, which were taken from prior studies (as shown in Table 2).Furthermore, we had three researchers (not the authors) with experience in internet marketing and sales test all of the measurements.We refined several items based on their comments.The revised questionnaire was then sent to 30 live-streaming video viewers to identify any remaining unclear issues.Following this procedure, a few small changes were made to improve the accuracy of the construction measurements.Because we carried out a survey in Vietnam, the translation-back-translation method was used to assure that the Vietnamese questionnaire was consistent with the original English version.
We selected Facebook Live, one of the most popular live-streaming shopping platforms in Vietnam.Participants have a wide range of experiences viewing live streams.This online questionnaire was shared on Facebook and other social media sites.We also sent this link to users who were viewing Facebook live streams.A filter question was included to eliminate those who had never watched live streaming before.In total, we collected 455 responses.By discarding the invalid responses such as missing answers, inconsistent answers, and identical answers, we had 360 responses (accounting for 79.12%) for further analysis by Partial Least Squares (PLS).As shown in Table 1, the gender was skewed toward females, which accounted for 78.3% of the sample.The majority of the respondents were under 22 years old, accounting for roughly 54.4%.Correspondingly, in this research, most of the respondents were students (55.8%) and had a monthly income of under 10 million VND (69.2%).In short, the sample is rather unbalanced in terms of age, occupation, and income because the remarkable characteristic of the live-streaming users is that they are young.

3-2-Analysis
Partial least squares-structural equation modeling was used applying SmartPLS 3.0 to evaluate the measurement model and structural model [43].First, the measurement model (the outer model) was assessed by examining indicator reliability (outer loadings), internal consistency (Cronbach alpha and composite reliability), convergent validity (average variance extracted -AVE), and discriminant validity (Fornell-Larcker criterion).Second, structural model assessment in PLS-SEM focuses on evaluating the significance and relevance of path coefficients, followed by the model's explanatory power (explained variance -R 2 ; effect size -f 2 ) and predictive relevance (Q 2 ).

4-Results and Discussions
We carefully tested for indicator and construct reliability as well as convergent and discriminant validity [43].Tables 2 and 3 illustrate the outcomes of the measurement model.All of the outside loadings are above 0.70, and Cronbach's alpha values are higher than 0.80.The CR values ranged from 0.917 to 0.937, which was significantly higher than the threshold of 0.70.Furthermore, the AVE values are from 0.712 to 0.828, exceeding the 0.50 cutoff.In addition, the Fornell and Larcker matrix was satisfied when the square root of the construct's AVE was higher than the correlation with any other construct.We also test for multicollinearity issues.All VIFs are less than 5, indicating that multicollinearity is not problematic in this study.As a result, we proceed to the following analysis: After confirming the measurement model's validity and reliability, SmartPLS 3.0 was used to test the hypothesized relationships.Figure 3 and Table 4 depict and present the results.The results show that value similarity, other members' endorsement, utilitarian value, and hedonic value positively impact para-social interaction, supporting H1, H2, H3, and H4.Next, utilitarian value, hedonic value, product knowledge, and para-social interaction positively impact trust in streamers, which supports H5, H6, H7, and H8.Besides, the effect sizes (f 2 ) of 0.02, 0.15, and 0.35 were used to describe small, medium, and large effects for exogenous latent variables, respectively [50].All of them were greater than 0.02, indicating small to medium effects.Finally, Table 5 indicates that the model can explain 72.1% of the variance in parasocial interaction and 65.4% of the variance in trust towards streamers.Besides, the blindfolding procedure with calculated Q 2 values greater than 0 supports the predictive capability of the research model.Table 4 shows that similarity value has a positive impact on para-social interaction (β = 0.206, P-value <0.01).This finding is in line with [21], indicating that similarity value plays an important role, so it should be taken into more consideration because consumers can easily find members with the same interests or tastes concerning products or services, lifestyle, or shopping experiences.When customers have the same value as live streamers, they tend to feel closer and engage in more para-social interactions with the live stream seller.
Next, other members' endorsements have a significant impact on para-social interaction (β = 0.204, P-value <0.01).The finding is consistent with [24,26], suggesting that positive endorsement can result in some great influences on parasocial interaction with streamers.It implies that the role of other members' activities, such as pressing the "like" sign on Facebook, commenting positively, or recommending, should be taken into more consideration because two main reasons.Firstly, word of mouth is not only economical but also highly effective in Vietnam's market.Secondly, Vietnamese customers usually choose live-streaming videos that are offered by their family, colleagues, or friends or have many viewers and positive comments.
Utilitarian value is presented to have a positive influence on para-social interaction (β = 0.172, P-value <0.01), supporting the finding of Lee et al. (2010) [30].First, Vietnamese consumers are often concerned with the reputation of streamers or online sellers.They tend to follow and watch video live streams of famous people like singers, artists, hot girls, superstars, successful people, or online sellers who have thousands of followers.Therefore, many businesses have used celebrities and social media influencers to market their products on live streams.Second, consumers like shopping on live streaming than on websites or blogs because many steamers have tried the product to display the quality and get instructions for use, making it easier for customers to visualize, assess, and make decisions.Third, clients always want to quickly receive responsiveness, which means that sellers need to respond immediately and quickly to customer requests, and concerns.
Hedonic value leads to positive para-social interaction (β = 0.442, P-value <0.01).The finding is in line with [40,51].It implies that Vietnamese consumers will choose video live streams that can make them feel pleasing and enjoyable to watch.Moreover, positive emotions and feelings can pave the way for emotional engagement with the seller.Following the report in the Facebook Video Summit 2021, watching live stream videos can help customers interact with sellers in live streams more easily than usual, about 1.3 times, and about 75% of viewers in live streams can make some actions, such as sharing, liking, and commenting with their family, friends, other persons, etc.
Similar to [9,37], utilitarian and hedonic values have strong impacts on trust (β = 0.283, p < 0.01; β = 0.162, p = 0.002), suggesting that online sellers should pay sufficient attention to utilitarian value and the experiential aspects of customers.First, streamers should inform shoppers about the benefits of shopping via live streams, such as increased visibility and faster response.Second, because watching live videos takes longer than browsing still images of things, live streamers should keep customers interested and prevent boredom by incorporating fun and exciting activities connected to products or incentives (e.g., games, gifts), as well as social interaction.The values can generate positive emotions, which can lead to emotional trust in streamers.
This study also discovered that para-social interaction during live streaming had a beneficial influence on viewer trust (β = 0.251, p < 0.001).As a result, live streamers should engage in more activities that encourage viewers to participate.Streamers can deliver customized demos and respond to consumers' queries more quickly via live streaming, which can decrease shoppers' doubt and boost their belief [37].During live streaming, they may recommend products based on individual preferences, as well as invite viewers to comment and vote, which has a direct impact on customer trust and engagement.Sellers should also take advantage of its visual elements to better describe their products and answer client inquiries.Facebook or social commerce designers could create funnier icons for people to interact with, increasing the satisfaction of interaction and, as a result, increasing sellers' trust.
The study once again confirms the findings of Chen et al. (2020) [37] that product knowledge of streamers had a positive effect on trust in the streamers (β = 0.270, p < 0.001).Before displaying products, live streamers should pay attention to their skills and learn more about them.When a streamer exhibits expertise, clients are more likely to trust the streamer.Before delivering advice to customers, live streamers should verify that they have a solid understanding of the product.Customers are more inclined to trust a streamer who proactively assists them in resolving a problem.

5-Conclusion
Our study adds to our understanding of commerce through live streaming by studying the impact of social interaction on trust in streamers.The findings could help streamers improve their social interactions with Facebook users, increasing customer trust in streamers.Streamers should use more attractive methods, develop more interesting activities, stimulate more comments, and increase engagement and connection.Para-social interaction should be directly tied to perceived similarity with a streamer in terms of shared value.As a result, it is a good idea to communicate trending ideals with a broad audience to both create social interactions with followers and increase their trust.Expertise and goodwill, which are essential for building credibility and persuasion, must be communicated through the developed content.
Several limitations should be discussed in this research.First, we exclusively collected data on live streaming on Facebook, even though there are other social media sites in Vietnam.As a result, the generalization of the findings to other social commerce platforms would be limited.Future research might incorporate and compare the samples from different nations to investigate potential differences in customers' trust.Second, the use of nonprobability survey data restricts the conclusions' generalizability.Future studies might use a longitudinal method to analyze customers' behavior in greater depth or investigate the moderating effect of gender on para-social interaction and trust in streamers.

6-2-Data Availability Statement
The data presented in this study are available on request from the corresponding author.

Figure 3 .
Figure 3.The results of the structural model Conceptualization, A.T.T. and T.D.P.M.; methodology, A.T.T, T.D.P.M., and T.H.M.T.; writing-original draft preparation, A.T.T, T.D.P.M., and T.H.M.T.; writing-review and editing, T.D.P.M., T.T.N, and T.T.T.L.All authors have read and agreed to the published version of the manuscript.