INTRODUCTION
Research background
Live streaming has emerged as a significant trend in the digital age, allowing users to easily engage with platforms for activities like gaming, singing, and selling from anywhere at any time During the COVID-19 pandemic, companies like Lazada and Shopee leveraged live streaming to boost sales, resulting in a 21-fold increase in viewership and a 24-fold rise in purchases via live streams compared to the previous year, according to Brands Vietnam Furthermore, the Vietnam E-commerce Association predicts that the rapid growth of e-commerce will continue at over 30%, with the market potentially reaching a value of $15 billion.
Live streaming distinguishes itself from other social media platforms by integrating features such as video content, consumption, and real-time communication (Lie et al., 2018) This innovative approach has transformed e-commerce into a social, hedonic, and customer-centered experience, moving away from the traditional product-oriented shopping model (Busalim, 2016; Wongkitrungrueng et al., 2018) As a novel business model, live-streaming commerce offers various stimuli that engage consumers and enhance their shopping experience.
Video streaming platforms such as Twitch, Facebook, YouTube, and Instagram have gained immense popularity, with Facebook emerging as the largest live streaming site globally, boasting 2.5 billion active monthly users as of Q4 2019, according to Restream However, a 2020 Vimeo report indicates that 70% of viewers prefer live streaming on YouTube Notably, China represents a significant market for live streaming, with the China Internet Network Information Center reporting 560 million live streaming users in March 2020, an increase of 163 million since the end of 2018, which constitutes 62% of the country's total internet users Additionally, the live streaming market in China reached a valuation of $16.3 billion in 2020, as highlighted in a Statista market research report.
Vietnam presents a significant opportunity for live-streaming commerce, with 62 million active social media users and 58 million mobile social media users, representing 64% of the population, according to Nielsen Vietnam (2019) Users spend an average of 2 hours and 31 minutes daily on broadcasting, streaming, and video on demand, with 99% of internet users engaging with online videos Tran Tuan Anh, managing director of Shopee Vietnam, highlighted that many brands and sellers view Shopee Live as an essential tool for adapting to changing consumer demands and effectively promoting their products Notably, there was a 70% increase in the total duration of Shopee Livestream from February to April 2020.
Live-streaming commerce is gaining popularity, yet it remains under-researched Existing literature primarily examines customer engagement, viewing intentions, and behavioral drivers, but lacks a comprehensive analysis of how contextual cues influence customer behavior within the live-streaming environment (Wongkitrungrueng et al., 2018) Additionally, the factors currently affecting customer behavior may differ from those identified in previous studies, highlighting the need for further exploration in this evolving context.
Live streaming is a powerful tool for attracting consumers and boosting product sales, particularly in the fashion industry Understanding the factors that influence customer intentions in live streaming shopping is crucial This research aims to explore the live streaming shopping intentions of Vietnamese consumers, providing a detailed perspective on their behaviors and preferences in this emerging market.
Research objectives
Derived from the context described above, this present research aims to contribute prior studies on identifying factors affecting live-streaming shopping intention in Vietnam The case of fashion products
In particular, the present study examines the proposed framework from previous theoretical studies to understand determinants on consumer‟s live streaming purchasing intention based on SOR theory
Based on the objective, I formulated two questions to conduct the research:
- What factors affect Live-streaming purchasing intention of consumers?
- How do these factors impact consumer shopping intention?
Research scope
Content scope: Factors affecting live streaming shopping intention in Vietnam The case of fashion products
Place scope: all the locations in Vietnam
Time scope: October 2020 to May 2021.
Research structure
The study has 5 chapters, including:
The first chapter introduces the research background which is the circumstance as motivation for this research to be conducted
The second chapter focuses on the literature review, offering essential definitions and summarizing prior research on consumers' intentions to purchase through live streaming It identifies existing literature gaps that serve as a basis for formulating hypotheses.
The third chapter is about research methodology, design and procedure, pilot test, survey adjustment, variables measurement, data collection and analysis method
The fourth chapter is data analysis and results
The final chapter is a discussion on findings, limitations of this research, the recommendation for future studies and implications if there is any.
LITERATURE REVIEW
Related Definitions
Live-streaming has become a significant trend in the economy, with live-stream selling emerging as a crucial skill for both individual sellers and e-commerce platforms Research by Smith et al (2013) highlights that live video streaming differs from other social media formats due to the presence of broadcasters or streamers In China, Lisa Magloff (2020) describes Taobao's live commerce as a platform where hosts showcase products and engage with users through live chats Similarly, studies by Bründl et al (2017) and Deshpande & Hwang (2001) characterize live-streaming commerce as a form of synchronous communication, allowing viewers to observe a seller's verbal and nonverbal behaviors This interactive format enables streamers to connect with multiple customers simultaneously, who can respond via text, making live streaming commerce a blend of real-time video and e-commerce (Wang, 2017) that fosters social interaction (Cai & Wohn, 2019).
Streaming services have emerged as a popular alternative to traditional broadcasting due to their superior quality and diverse content offerings (Singh et al., 2020) Live-streaming platforms allow users to generate their own content—ranging from gaming and cooking to singing and beauty vlogging—creating a dynamic interaction with followers (Recktenwald, 2017) As noted by Rhea Liu and Dannie Li (2016), live-streaming serves as a "time killer," providing viewers with companionship and a form of interaction that mimics real-life communication, offering a relaxing escape from everyday stress.
According to Apiradee et al (2020), live streaming commerce can occur through three main channels: dedicated live streaming platforms that incorporate commercial activities (like TikTok), e-commerce marketplaces and mobile apps that integrate live streaming features (such as Shopee), and social networking sites (SNSs) that enhance their platforms with live-streaming capabilities (like Facebook Live) to promote sales This research will specifically examine e-commerce marketplaces and social media networks with live-streaming functions to analyze viewers' purchasing intentions regarding fashionable products.
2.1.2 Stimulus-Organisim-Respone (SOR) Model
The S-O-R (Stimulus-Organism-Response) model, developed by Mehrabian and Russell in 1974, emphasizes how environmental stimuli influence individual emotional experiences and subsequent behaviors, making it a key framework in retail contexts such as purchasing decisions, impulse buying, and self-service interactions Research has demonstrated the model's effectiveness in linking emotional responses to consumer actions, including intention and purchase behavior Recently, scholars have applied the S-O-R framework to online consumer behavior, examining factors like trust and repurchase intentions, as well as the impact of online atmospheres on consumer interactions In the realm of live streaming commerce, the S-O-R model has been utilized to analyze various environmental stimuli, such as website content, streamer appeal, and social influences, highlighting its capacity to assess the emotional and cognitive responses of consumers Given the unique convenience of streaming services, traditional models like TAM and UTAUT fall short in explaining viewer intentions, underscoring the need for the S-O-R framework to explore the interplay between contextual cues and consumer decision-making processes.
Mei Teh Goi et al (2014), Daunt and Harris (2012), Lin (2004), and Wong et al
(2012) indicated that stimulus directly influence customers‟ response
Three major determinants of the S-O-R model usually demonstrate in a variety of dimensions; nevertheless, within the live streaming context, these elements are specified in this study as follow:
Stimulus (S) refers to the elements that engage consumers, including streamer attractiveness, information quality, and social interaction (Chan et al., 2017) In the context of live streaming, streamers play a vital role, as they generate engaging content, provide valuable product information, and facilitate real-time interactions with consumers during broadcasts.
- Organism (O), which is an internal evaluation of consumers (Chan et al., 2017): live-streaming trust
- Response (R), which is an outcome of consumers‟ reaction(s) toward the online shopping stimuli and their internal evaluations (Chan et al., 2017): live-streaming shopping intention
There have many studies indicating the effect of attractiveness on tangible benefits
Research indicates a significant correlation between physical attractiveness and various factors such as job offers, future success, and affiliation with high-status social groups (Michael et al., 2009; Vicki et al., 1992; Anne et al., 2011) Interpersonally attractive individuals are often viewed as credible sources and tend to receive more favorable evaluations Wohn et al (2018) found that traits like interpersonal and physical attractiveness in streamers are linked to increased emotional support from viewers, which encompasses affection and encouragement Additionally, Xu et al (2020) characterized streamers as "endorsers" in live streaming commerce, while Baker and Churchill (1977) noted that attractive endorsers effectively influence consumer attitudes and beliefs about products Similarly, Frevert & Walker (2014) highlighted that attractive individuals are perceived as more popular and receive higher evaluations compared to their less attractive counterparts.
Research by Fang et al (2020) highlights the physical attractiveness stereotype, linking attractive individuals to positive personality traits such as warmth, kindness, trustworthiness, and sociability Additionally, Zhao et al (2015) conducted experiments showing that facial attractiveness influences both implicit and explicit trusting behaviors Their findings suggest that businesses can benefit from presenting attractive images and sellers, as consumers who perceive beauty in their surroundings are more likely to feel comfortable and in a positive mood, leading to increased trust in products and a higher intention to purchase.
While the impact of streamer attractiveness on consumer trust in live-streaming commerce remains underexplored, it is essential for understanding purchasing intentions in this industry Research suggests that attractive streamers are more likely to foster trust among viewers, leading to increased consumer engagement Therefore, we propose the first working hypothesis: attractive streamers will significantly enhance consumer trust, ultimately influencing purchasing decisions in live-streaming contexts.
H1 Streamer attractiveness is positively associated with trust
Research by Nhu-Ty Nguyen (2021) highlights that the physical attractiveness of celebrities significantly influences the purchasing intentions of young Vietnamese consumers Similarly, Juulia Karaila (2021) emphasizes the crucial role of social media influencers' attractiveness in positively affecting consumers' purchase decisions.
H2 Streamer attractiveness is positively associated with Live streaming shopping intention
In the realm of online shopping, consumers rely heavily on digital information such as images, videos, and product descriptions since they cannot physically interact with the items According to Wolfinbarger and Gilly (2001), the information provided by online sellers must be relevant and assist consumers in assessing the quality and usefulness of products or services Research by Wang and Strong (1996) and Zhang et al (2000) emphasizes that effective information should be current, comprehensive enough to aid decision-making, consistent in presentation, and easy to comprehend.
High-quality information plays a crucial role in enhancing consumer buying decisions and satisfaction, as highlighted by Peterson et al (1997) and further developed by Park and Kim (2003), who noted that both product and service information quality significantly influence information satisfaction This quality is essential for reducing transaction costs, minimizing perceived risks, boosting consumer confidence, and improving the overall shopping experience (Gao et al., 2012; Nicolaou et al., 2013) According to the 2015 Shotfarm Product Information Report, 78% of consumers consider product content quality a critical factor in their purchasing decisions, with one in four abandoning purchases due to inadequate product information Therefore, understanding the impact of information quality is vital for discussing factors that affect consumer purchase intentions in both live-streaming commerce and e-commerce.
Cyr (2008) identified that the design of information significantly affects trust, particularly regarding the accuracy of product details on e-commerce platforms In the context of live-streaming commerce, consumers benefit from high-quality information through various cues such as images, review comments, audio from sellers, detailed product demonstrations, and real-time interactions These elements allow consumers to vividly understand product functionality in live-stream videos Consequently, the quality of information presented can lead viewers to reassess their perceptions of products (Xu et al., 2020), ultimately enhancing viewer trust.
H3 Information quality is positively associated with trust
There was many scholars mentioned the relationship of information quality and purchase intention in their studies Chiu, Hsieh and Kao (2005) suggest that information quality is related to the behavioral intention of customers in terms of intention to use the website to purchase, intention to recommend it to other people), what was also verified by Kim and Niehm (2009) Furthermore, to support for these oppinions, G S Milan et al.,
(2016) identified information quality as an antecedents of purchase intention Thus, the fourth research hypothesis emerges:
H4 Information quality is positively associated with Live streaming shopping intention
In the context of Live-streaming commerce, Interactivity is a key characteristic, which fosters viewers‟ attitudes and behaviors in communications and transactions There are many different perspectives in literature defined about interactivity; however, in the study, I agree that interactivity refers to the degree to which interactions occur in mutual communication between two parties (Kang et al., 2021; Bonner, 2010; Lee, 2005)
Interactivity is shown at a high-quality communication in live-streaming compared with other e-commerce forms because live-stream shopping platforms are considered as a unique form of social media that help users to interact with streamers as well as with other viewers (Zhao et al., 2018) In other words, viewers could share their thoughts and messages in real-time; while, streamers would react, respond and feedback immediately to audiences‟ requirements/questions/comments by talking in the live-stream or performing certain activities Similarly, users might interact with co-viewers by chatting, following and debating other‟s comments This allows viewers to be perceived the useful information and the care of streamers about what they expect and act, which can motivate and enhance trust on sellers or streamers of participants in a live-streaming video
Research conceptual model
Research model is depicted and developed based on the explaination of the relationship between the identified dimensions (Figure 2.1) SOR should be applied in the research because the framework has been used widely in various psychology researches to study consumer behaviors Furthermore, many recent scholars approached SOR theory in their research and gain deeper knowledge in Live-streaming context Therefore, the research model relied on SOR theory posits that three stimulus including streamer attractiveness, information quality and interactivity affect trust, resulting in purchasing intention in Live-streaming commerce I also determine that trust mediate the relationship between three stimuli and the response
A number of studies have addressed consumer behavior in the context of Live streaming commerce ( Xu et al, 2020; Chen et al, 2020; Sun et al, 2019; Venkatesh et al, 2012; Liu, 2003;) Nevertheless, these studies approach mainly the China culture where the development of Live streaming has been very modern Meanwhile, the problem in Live streaming shopping intention of Vietnam, in particular, is different and there have been not many Vietnamese studies researched this field Some Vietnamese articles researched Live streaming as a business model for the teaching or education sector, others mentioned consumer‟s buying intention in Live streaming but it focused on only Facebook platforms
In addition, Trust is very important in both online and offline environment which was confirmed by various studies Moreover, some scholars demonstrated that trust is a factor of organisms
Chen and Barns (2007); Winch and Joyce (2006); Dash and Saji
Trust is very important in the process of decision making when consumers purchase a product in both offline and online environments
Building trust may help develop business transactions or some responses such as purchase intention in an online environment
Show the significant mediating role of trust with online purchase intention
B Zhu et al., (2019), Linlin Zhu et al., (2020)
Demonstrate that trust is a factor of organisms (O)
Despite the growing interest in live streaming, there is a notable deficiency in comprehensive research regarding trust within this context Recent studies have utilized the SOR model to investigate customer behaviors in live streaming, yet they have overlooked the critical aspect of trust in their analyses.
My research aims to explore whether trust is an internal state that influences customer purchasing intentions in the context of live streaming, utilizing the S-O-R (Stimulus-Organism-Response) theory to bridge this gap.
RESEARCH METHODOLOGY
Research process
This research was conducted following the steps shown in the figure below:
Figure 3.1 Research process proposed by the author
Review the literature and related researches
Identify research model, hypotheses and methodology
Identify research population,sample, scale and measurements
Plan to do survey and develop questionnaire based on previous studies
Meeting with supervisors for finalizing the plan for survey and questionnaires Conclusions and suggestions
Questionnaire Construction
The research utilized instruments developed from prior literature to ensure validity across five key constructs: streamer attractiveness, information quality, interactivity, trust, and purchase intention Each construct was assessed using a five-point Likert scale, ranging from "strongly disagree" (1) to "strongly agree" (5).
The Likert scale, developed by Rensis Likert in 1932, is widely utilized in survey research due to its simplicity and ease of use According to Neuman (2000), the data obtained from this scale is highly regarded, prompting the author to employ this method for creating a questionnaire survey.
Totally Disagree Disagree Consider/Normal Agree Totally Agree
The survey was divided into two sections: the first part included general demographic questions regarding participants' gender, age, educational background, and their experiences with live-streaming commerce, such as frequency and duration of viewing The second part featured 22 closed-ended questions designed to assess the importance of factors like streamer attractiveness, quality of information, interactivity, and trust in influencing shopping intentions in live-streaming environments.
The quality of information was assessed through five key dimensions: accuracy, timeliness, adequacy, completeness, and credibility, as outlined by Chen et al (2020) Respondents evaluated their trust in the information based on these dimensions using a seven-point Likert scale, where (1) represented "not timely" or "not accurate" and (5) indicated "very timely" or "very accurate."
The level of consumer trust was measured using Wongkitrungrueng & Assarut,
(2020) This measure reflects the reliability of the buyer to the streamers in Live streaming shopping
The quality of information significantly influences trust, which consists of expectations and concern Higher information quality—characterized by timeliness, accuracy, adequacy, completeness, and credibility—enhances trust levels among consumers When streamers deliver such information, it demonstrates their professionalism, leading consumers to appreciate their skills and expertise This clarity in expectations fosters a trusting relationship, as consumers are more likely to engage in purchasing behaviors that reflect their desire for streamers to effectively meet their needs Ultimately, both aspects of trust are fulfilled through effective communication and professional performance.
Constructs Item Scales Scales reference
SA1 I think that the live stream streamer is talented
SA2 I think the live streaming style of streamer is enjoyable
SA3 I think that the streamer has an interesting personality
SA4 I think the streamer has an appealing appearance
IQ1 I think the content provided by the streamer is reliable (such as product, brand, and use experience) Xu et al., (2020)
IQ2 The streamer provides real-time information to meet my needs in the live stream
IQ3 The streamer provides in- dept./detailed information about the fashion products (materials, colors, )
IQ4 The streamer provides accurate information about the fashion product that I want to purchase
IQ5 The streamer provides up-to-date information about the fashion product in live streaming video
I1 When watching a live-stream, I can exchange and share opinions with the streamer or other audiences easily
I2 When watching a live-stream, I feel closer to the streamer
I3 I feel that streamers care my respond in live streaming
I4 When I am watching a live-stream, the streamer provides sufficient opportunities to talk and ask a question
I5 The streamers respond to my question very fast
T1 I believe in the information that the streamer provides through live streaming
T2 The sellers in live-streaming commerce are trustworthy
T3 I think fashion products I order from Live streaming will be as I imagined
T4 I trust that the products I receive will be the same as those shown on
LSI1 I will continue watching Live streaming for purchasing fashion products
LSI2 I will consider live streaming shopping as my first shopping choice
Sun et al., (2019) LSI3 I expect that I will purchase fashion products through live streaming shopping in the near future
LSI4 I will recommend people around me to watch Live-streaming for purchasing fashion products
Sample and data collection
Sample size has to large enough to analyze factors, according to Hair et al., (1998), sample size (n) has to equal 5 times observed variables (m) or Tabacknick và Fidell
(1996) said that sample size n >= 50 + 8*m (n: sample size; m: independent factors)
Collecting an adequate number of respondents is crucial for applying Structural Equation Modeling (SEM) to analyze primary data According to Hair et al (1998), sample size significantly influences the estimation and interpretation of SEM results Scholars generally agree that larger samples yield more stable parameter estimates, with Bollen (1989) and others suggesting a minimum of 200 observations as a standard for research However, Hair et al (2014) indicated that there is no definitive rule for sample size, and it can be increased if technical issues arise in the research model.
To make sure the representation of the study, the author predicted the sample size was 200 and the actual respondents collected was 349 which was suitable with the research objectives
In this research, there are two ways used to collect data including primary data and secondary data
The author conducted an online survey to gather primary data on live-streaming shopping intentions among consumers in Vietnam from January to April 2021 A total of 241 questionnaires were collected through an online platform, with additional printed versions distributed at Thuong Mai University and Minh Khai High School in Hanoi Out of 349 responses received, 332 were deemed valid, as 17 were discarded due to random or inattentive answers Participants included individuals with experience in watching live-streaming videos or purchasing fashion products through live-streaming, and they were asked to respond based on their intentions for live-streaming commerce.
Secondary data refers to information gathered from previous studies, scientific journals, and relevant articles that align with specific research objectives and issues This data is sourced from credible internet platforms, including company websites and official sources, and serves as a valuable reference for research purposes.
Data analysis
The data for this study was collected through a survey, where unqualified respondents were first identified and removed The validated primary data was then entered into Excel for storage and subsequently transferred to SPSS for descriptive statistical analysis Concurrently, SMARTPLS 3.0 was utilized to assess measurement models, evaluate the significance of various factors, and test the proposed hypotheses.
3.4.1 Cronbach‟s Alpha & Explore factor analysis
The research evaluated the reliability of Cronbach's Alpha measurement, indicating a strong correlation among observed variables within the same factor This analysis identifies which variables contribute to the measurement of a factor's concept and which do not For reliable results, the Cronbach's Alpha Coefficient must exceed 0.6, and the Corrected Item-Total Correlation should be greater than 0.3, as noted by Hafiz and Shaari (2013).
Exploratory Factor Analysis (EFA) is a crucial component of quantitative analysis using SPSS, as it examines the relationships among variables across various groups or factors This method helps identify observed variables that load onto multiple factors or those that may have been incorrectly assigned to factors initially To effectively analyze EFA, certain criteria must be considered.
- Kaiser-Meyer-Olkin (KMO) is the is an indicator used to consider the appropriateness of EFA The value of KMO must be higher than 0.5 (0.5 ≤ KMO
≤ 1) If this number is less than 0.5, factor analysis is likely not suitable for collected data
- Bartlett‟s test of sphericity is used to examine observed variables have correlation or not The condition of Bartlett‟s test is that Sig Bartlett‟s Test < 0.05
In exploratory factor analysis (EFA), a Total Variance Explained greater than 50% indicates its suitability, revealing the percentage of variance accounted for by the extracted factors and the percentage of observed variables that are not captured (Hair et al., 2006).
Factor loading reflects the correlation between observed variables and underlying factors, with higher values indicating stronger relationships According to Hair et al (2009), a factor loading of 0.5 is considered significant for sample sizes between 120 and 350, while a loading of 0.3 is deemed acceptable for larger samples exceeding 350 (Hair et al., 2006).
Over 20 years, many scholars have applied structural equation modeling (SEM) in their research as the second generation data analysis techniques According to Byrne
Structural Equation Modeling (SEM) is a collection of multivariate techniques that focuses on confirmatory analysis to assess model fit with data, rather than exploratory analysis Unlike multivariate regression, SEM enables the simultaneous examination of multiple relationships, investigating interrelated dependence among measured variables and latent constructs, as well as the connections between various latent constructs (Hair et al., 2011).
Structural Equation Modeling (SEM) consists of two key components: the measurement model and the structural model The measurement model is essential for defining and validating the reliability and validity of constructs, as outlined by Hair et al (2011) It also estimates the statistical significance of path coefficients and hypothesis testing within the structural model A structural relationship can only be established once the measurement model demonstrates an acceptable fit.
The evaluation of the measurement model using PLS algorithms in SMARTPLS is crucial for assessing measurement reliability Key indicators to focus on include outer loading, reliability, convergent validity, and discriminant validity.
Outer loading The degree of association ≥ 0.708: quality observed between the observed variable with the latent variable variables [Hair et al., (2016)]
- Cronbach's Alpha (CA) ≥ 0.7 [(De Vellis R.F., (2012)]
- Composite Reliability (CR) ≥ 0.7 [Hair et al., (2010); Bagozzi
Convergent validity Average Variance Extracted AVE
- HTMT [Henseler et al., (2015)] the degree of distinguishing a concept of a particular latent variable from the concept of another latent variable [Henseler et al., (2009)]
The evaluation of the structural model assesses the relationships between concepts, focusing on how independent variables influence a dependent variable through an intermediate variable Key criteria for this evaluation include analyzing the impact and intensity of these relationships.
The R-square value is a crucial metric that indicates how well data aligns with a given model, reflecting the model's explanatory power According to Henseler et al (2009), in Partial Least Squares (PLS) path models, R-square values of 0.67, 0.33, and 0.19 are categorized as strong, medium, and weak, respectively.
The path coefficients in a structural model must be assessed based on their sign, magnitude, and significance, similar to standardized beta coefficients in ordinary least squares regressions To establish confidence intervals for these path coefficients and conduct statistical inference, bootstrapping techniques should be employed (Henseler et al., 2009).
- T-statistics are generated to assess the significant level of the measurement model and structural model T-statistics is greater than 1.96 which indicates a statistical significance of hypotheses tested [Hair et al., (2012)].
DATA ANALYSIS
Measurement Model Test
SPSS and Smart PLS were used to test the research model and hypotheses
At least once per day
Less than once per week
Watch duration Less than 30 minutes each time
No experience/Less than 1 month
The total of valid respondents who participated in the research is 332, of which
Out of 332 participants, 229 are female (68.98%) and 103 are male (31.02%), indicating an unequal gender distribution in the responses The majority of respondents fall within the 25 to 30 age range, comprising 44.78% of the total, while only 1.5% are over 40 years old Additionally, 22.39% of participants are aged 30 to 35 Overall, young individuals aged 20 to 34 make up 70% of the participants.
The majority of participants engaged in live streaming less than once a week (33.73%), with watch durations typically ranging from one to two hours (43.67%) Most participants reported having 4-12 months of experience purchasing products on online platforms These statistics indicate that the participants are well-qualified to provide valuable insights for our research, ensuring the reliability of the results.
Cronbach‟s Alpha measurement
The reliability of the instrument was assessed using Cronbach's Alpha, which revealed values exceeding 0.6 for all constructs, including streamer attractiveness, information quality, interactivity, trust, and live streaming purchase intention Notably, all measured variables maintained a correlation above 0.3, with the exception of IQ4, which was deemed an unvalued variable and subsequently excluded from the model Consequently, 21 observed items were retained for further analysis through Exploratory Factor Analysis (EFA).
Table 4.1 Cronbach’s Alpha result 1 (Analyzed by SPSS)
Variables Cronbach's Alpha Item‟s quantity
IQ (Information quality) 0.787 4 (removed IQ4)
LSI (Live streaming purchasing intention) 0.796 4
Exploratory Factor Analysis (EFA)
4.3.1 Exploratory factor analysis of Stimulus scale
The initial stimulus scale consisted of 13 variables, with a KMO value between 0.5 and 1 indicating acceptable measurement quality (Hair et al., 1998) Bartlett's test of sphericity showed a significant correlation among the observed variables, with a coefficient of Sig = 0.000 To ensure valid results, the Total Variance Explained must exceed 50% (Gerbing and Anderson, 1988), and factor loadings should be above 0.5 for reliable analysis Following a reliability analysis, one item was removed due to insufficient reliability, leaving 12 observed variables that demonstrated internal consistency An exploratory factor analysis was then performed on these 12 variables to assess their convergence with the components.
The second round of Exploratory Factor Analysis (EFA) yielded a KMO value of 0.905, with Bartlett's Test showing significant results (Sig = 0.000), indicating that the observed items were correlated within the dataset Due to a factor loading of I4 being below 0.5 in the initial analysis, I4 was removed from the model The cumulative sum of squared loadings reached 54.817%, surpassing the 50% threshold, which signifies that 54.817% of the variance in the stimuli is accounted for by three factors.
Table 4.2 EFA results of Stimulus Scale
Kaiser-Meyer-Olkin Measure of Sampling Adequacy
Rotation Sums of Squared Loadings a Total
4.3.2 Exploratory factor analysis of Organism scale
The organism scale consists of four items, and the KMO and Bartlett's Test reveals a KMO value of 0.765, indicating adequate sampling, as it exceeds the 0.5 threshold With a significance level of 000, these results validate the data for exploratory factor analysis.
Table 4.3 EFA results for Organism scale
Kaiser-Meyer-Olkin Measure of Sampling
Initial Eigenvalues Extraction Sums of Squared Loadings
Eigenvalues is more than 1 (= 2.677) with cumulative variance = 55.958% explaining 55.958 % of trust
In short, after analyzing exploratory factor, the organism scale remains with 4 variables, extracted to 1 component – Trust
4.3.3 Exploratory factor analysis of the response scale
Table 4.4 EFA results for Organism scale
Kaiser-Meyer-Olkin Measure of Sampling Adequacy
Loadings Total % of Variance Cumulative % Total
Extraction Method: Principal Axis Factoring
The response scale consists of four items, with the KMO value measuring sampling adequacy at 0.696, which exceeds the threshold of 0.5 Additionally, the significance level is 000, confirming the data's validity for exploratory factor analysis.
Eigenvalues is more than 1 (= 2.509) with cumulative variance = 50.752% explaining 50.752% of Live streaming purchase intention
In short, after making exploratory factor analysis, the response scale remains 4 observed variables, extracted to 1 component – Live streaming purchase intention
After making exploratory factor analysis, there are no changes in the components of the model Therefore, research model is not adjusted and remained the same with the proposed hypothesis.
Structural Equation modeling (SEM)
Structural Equation modeling (SEM) was appropriated by Smart PLS 3.2.9
In the initial assessment, the outer loading of most constructs in the research exceeded 0.7, with the exception of LSI2, which necessitated its removal from the model Consequently, a second round of analysis was conducted, and the final results are presented in Table 4.5.
Look at table 4.5, the outer loading of all items was higher than 0.7 which means that 19 remain observed items were significant in the research model
Table 4.5 Outer Loading of the constructs
The study assessed measurement reliability using composite reliability (CR), Average Variance Explained (AVE), and outer loading The CR values exceeded the acceptable threshold of 0.7, ranging from 0.875 to 0.890, while the AVE values also surpassed the required 0.5, ranging from 0.637 to 0.669 These findings indicate satisfactory convergent validity.
Table 4.6 Construct Reliability and Validity
The research assessed the discriminant validity by calculating the square root of each Average Variance Extracted (AVE) The findings indicate that all constructs within the research model demonstrate acceptable discriminant validity, as the square root of each AVE exceeds the inter-construct correlations.
Table 4.7 Correlation among Constructs and AVE square root
The second method to examine the discriminant is Heterotrait-Monotrait Ratio (HTMT)
Look at this table (HTMT), all values HTMT were less than 0.85; thus the discriminant validity of all constructs could be ensured
Data analysis results indicate that trust accounts for 30.4% of the variance in live streaming intention (R² Adjusted = 0.304) Additionally, independent variables such as streamer attractiveness, information quality, and interactivity collectively explain 47.1% of trust (R² Adjusted = 0.471).
The results of the Structural Equation Modeling (SEM) indicate that the model explains 30.4% of the variance in purchase intention, while trust accounts for 47.1% of the variance Significant paths were identified through 1,000 bootstrap runs, revealing that streamer attractiveness, information quality, and interactivity positively influence trust, with coefficients of 0.471, 0.206, and 0.130 (p ≤ 0.005), respectively, thus supporting hypotheses H1, H3, and H5 Additionally, streamer attractiveness positively affects live streaming purchase intention (β = 0.128, p < 0.005), and trust significantly impacts purchase intention (β = 0.444, p < 0.005), confirming hypotheses H2 and H7 However, interactivity and information quality do not have a direct effect on live streaming shopping intention (p-value > 0.05), leading to the rejection of hypotheses H6 and H4.
Table 4.9 Mean, STDEV, T-Values, P-Values
CONCLUSION
Discussion on Findings
This study explores the factors that influence trust in live-streaming platforms and their impact on consumers' purchase intentions Utilizing the S-O-R framework, an empirical test was conducted to identify key determinants that foster consumer trust, particularly in the context of fashion products The findings reveal both theoretical and practical implications for enhancing trust and driving purchase intentions in live-streaming environments.
The study highlights the significant role of trust in live-streaming purchasing intentions, aligning with previous research on online trust Trust fully mediates the effects of information quality and interactivity on consumer purchase intentions, while also partially mediating the attractiveness of streamers Viewers are drawn to the streamers' appearance and communication style, fostering long-term trust in both the streamers and their products Additionally, streamers effectively present high-quality product information and respond to customer inquiries in real-time, enhancing understanding and visualization The engaging nature of live videos, combined with the appeal of attractive streamers, generates positive emotions and builds affective trust in both the products and the sellers, ultimately leading to a higher likelihood of purchase.
Research indicates that factors such as streamer attractiveness, information quality, and interactivity significantly influence trust in live-streaming, particularly in the fashion sector Notably, streamer attractiveness has the most substantial effect on trust, surpassing other factors, despite previous studies suggesting its impact on purchasing intention is less pronounced Streamers enhance viewer engagement by sharing relatable stories and experiences, which not only entertain but also provide valuable fashion insights This connection fosters admiration and trust among viewers, leading them to support streamers through purchases, thereby establishing long-term relationships.
The study highlights that interactivity plays a crucial role in building consumer trust in live-streaming contexts, supporting hypothesis 5 This finding aligns with previous research indicating that increased interactivity fosters trust among online consumers Specifically, a higher level of interaction between viewers and streamers, as well as among co-viewers, enhances trust and boosts purchase intentions Live-streaming facilitates this engagement through real-time broadcasts and text chat features, allowing customers to pose questions and receive immediate responses from streamers Additionally, viewers can observe the inquiries and feedback from other participants, providing a wealth of credible information that aids in their purchasing decisions.
Research indicates that while information quality is a significant factor influencing trust in live-streaming commerce, its impact may be less pronounced than previously thought Xu et al (2020) highlighted that information quality is crucial for cognitive assimilation; however, consumers often lack sufficient information during live broadcasts This inadequacy can lead to confusion regarding product quality and associated risks Additionally, past experiences reveal that viewers who have purchased items through live-streaming often encounter unexpected product performance due to inaccurate information provided by streamers As a result, consumers may struggle to fully trust the information presented, undermining their confidence in the live-streaming shopping experience.
Contribution of Research
The study highlights important theoretical contributions by demonstrating the influence of contextual factors on consumer trust, which subsequently affects purchasing intentions in live streaming Notably, among the three stimuli examined, streamer attractiveness emerges as the most significant factor impacting trust in the live streaming environment.
Investigating trust as a key driver of consumer purchase intention reveals its significant impact on live streaming purchasing behavior Research indicates that consumer trust plays a crucial role in influencing purchasing intentions, as supported by various studies (e.g., Chang et al., 2015; Dabbous et al., 2020) The findings provide statistical evidence that trust positively affects live-streaming purchasing intention, enhancing our understanding of the organism's response within the SOR model framework.
The study highlights the crucial mediating role of trust in the relationship between streamer attractiveness, information quality, interactivity, and purchasing intention in live streaming commerce Although streamer attractiveness has a direct but minimal impact on purchase intention, trust significantly enhances consumers' likelihood to buy Unlike previous research that primarily focused on cognitive and emotional responses, this study emphasizes trust as a key factor influencing purchasing decisions in live streaming These findings deepen our understanding of the dynamics between stimuli, the organism (trust), and consumer responses in this emerging market.
Practical implication
This study provides valuable insights for streamers and live-streaming shop owners on enhancing strategies to attract consumers and maximize profits Building consumer trust is crucial, as a lack of trust in products or streamers can hinder live-streaming transactions To improve consumer trust, it is recommended that live-streaming shop owners focus on three key areas: streamer attractiveness, information quality, and interactivity, which can ultimately lead to increased product purchases.
To boost viewers' purchasing intentions, streamers and sellers should focus on enhancing their image and developing unique personal styles and skills Attractive and charismatic streamers who excel in communication can significantly engage audiences Additionally, live-streaming platform managers should consider recruiting appealing personalities, such as KOLs, celebrities, or social media influencers, to effectively promote and showcase products.
To enhance viewer engagement, streamers should prioritize interactivity by utilizing features such as chat boxes, donation options, and emoji buttons Beyond merely showcasing fashion products, they can foster connections by sharing personal anecdotes or humor Additionally, rewarding audiences with gifts or prompt replies encourages participation and idea-sharing This approach helps viewers perceive streamers as trustworthy friends, ultimately strengthening the relationship and increasing the likelihood of purchasing products during live-streaming sessions.
Enhancing information quality is crucial in live streaming commerce, as streamers must deliver accurate and reliable details about fashion products during broadcasts This includes providing comprehensive information such as brand names, colors, materials, usage methods, and key features of the products Additionally, ensuring clear and real-time visibility of product images through stable networking and well-designed streaming environments is essential Streamers should also tailor the information presented to meet consumer needs, incorporating feedback and addressing viewer questions effectively.
Limitations and Future Research Directions
This study has several limitations, primarily related to the data collection process, which was conducted online due to the complexities of the COVID-19 pandemic The respondents were exclusively from Vietnam, representing diverse backgrounds, which may not accurately reflect live-streaming purchasing intentions in other countries Future research should aim to include larger, more varied regional samples across different cultural and economic contexts, and comparisons of consumers' purchasing intentions on a national level should be undertaken to enhance understanding and yield more valuable insights into live-streaming commerce.
The current research focuses exclusively on fashion products, which may limit the applicability of the findings to other categories Different product types, such as technology devices, automotive components, and furniture, could yield distinct results regarding their influence on consumer trust and purchasing intentions during live-streaming broadcasts.
Finally, the research model could be extended by integrating additional factors or moderators including gender differences or personality traits of customers
Chan, T K H., Cheung, C M K., & Lee, Z W Y (2017) The state of online impulse-buying research: A literature analysis Information & Management, 54(2), 204–
The study by Chang et al (2015) explores how website quality and perceived trust influence customer purchase intentions in the hotel industry It highlights the roles of website branding and perceived value as moderating factors in this relationship The findings suggest that enhancing website quality and fostering trust can significantly boost customer confidence and willingness to make reservations, ultimately impacting hotel sales.
Chen, Z., Cenfetelli, R., & Benbasat, I (n.d.) The Influence of E-Commerce Live
Streaming on Lifestyle Fit Uncertainty and Online Purchase Intention of Experience Products 10
Chen, C.-C., & Lin, Y.-C (2018) What drives live-stream usage intention? The perspectives of flow, entertainment, social interaction, and endorsement Telematics and Informatics, 35(1), 293–303 https://doi.org/10.1016/j.tele.2017.12.003
Chen, Y.H and Barns, S (2007) „Initial trust and online behaviour‟, Industrial Management and Data Systems, Vol 107, No 1, pp.21–36
Chun-Der Chen, Qun Zhao & Jin-Long Wang (2020) How livestreaming increases product sales: role of trust transfer and elaboration likelihood model, Behaviour & Information Technology, DOI: 10.1080/0144929X.2020.1827457
Dash, S.B., and Saji, K.B.(2006) „Role of effective website-design in online shopping: a large scale empirical study in the Indian context‟, in Panda, T.K and Donthu,
N (Eds.): Marketing in the New Global Order, Challenges and Opportunities, Excel Books, pp.435–451
De Vellis R.F (2012) Scale Development: Theory and Applications Thousand
Fang, Y.H (2012) „Does online interactivity matter? Exploring the role of interactivity strategies in consumer decision making‟, Computers in Human Behavior, Vol 28, No 5, pp.1790–1804
Fang, S., Zhang, C., & Li, Y (2020) Physical attractiveness of service employees and customer engagement in tourism industry Annals of Tourism Research, 80, 102756 https://doi.org/10.1016/j.annals.2019.102756
Henseler, J., Ringle, C M., & Sinkovics, R R (2009) The use of partial least squares path modeling in international marketing In R R Sinkovics & P N Ghauri (Eds.), Advances in International Marketing (Vol 20, pp 277–319) Emerald Group
Publishing Limited https://doi.org/10.1108/S1474-7979(2009)0000020014
Hu, M., Zhang, M., & Wang, Y (2017) Why do audiences choose to keep watching on live video streaming platforms? An explanation of dual identification framework Computers in Human Behavior, 75, 594–606 https://doi.org/10.1016/j.chb.2017.06.006
Juulia Karaila (2021) The role of social media influencer characteristics on consumer behaviour, http://urn.fi/URN:NBN:fi:aalto-202105096552
Kang et al (2021) explore the impact of interactivity on customer engagement behaviors within live streaming commerce platforms, emphasizing the role of tie strength Their research, published in the International Journal of Information Management, highlights how stronger connections between users enhance engagement levels The findings underscore the importance of fostering interactive experiences to boost customer involvement in live streaming environments For more insights, refer to the article [here](https://doi.org/10.1016/j.ijinfomgt.2020.102251).
Kim, J.-H., & Park, J.-W (2019) The Effect of Airport Self-Service Characteristics on Passengers‟ Perceived Value, Satisfaction, and Behavioral Intention:
Based on the SOR Model Sustainability, 11(19), 5352 https://doi.org/10.3390/su11195352
Kim, Y., & Peterson, R A (2017) A Meta-analysis of Online Trust Relationships in E-commerce Journal of Interactive Marketing, 38, 44–54 https://doi.org/10.1016/j.intmar.2017.01.001
Leong, L.-Y., Hew, T.-S., Ooi, K.-B., & Chong, A Y.-L (2020) Predicting the antecedents of trust in social commerce – A hybrid structural equation modeling with neural network approach Journal of Business Research, 110, 24–40 https://doi.org/10.1016/j.jbusres.2019.11.056
Liao, S.-H., Chung, Y.-C., & Chang, W.-J (n.d.) Interactivity, engagement, trust, purchase intention and word-of-mouth: A moderated mediation study 22
Liu, C (n.d.) Exploring consumers’ purchase intention in social commerce 21
Milan, G.S., Bebber, S., Toni, D.D., & LucieneEberle (2016) Information Quality , Distrust and Perceived Risk as Antecedents of Purchase Intention in the Online Purchase Context.DOI: 10.15640/jmise.v2n2a2
NGUYEN, Nhu-Ty (2021) The Influence of Celebrity Endorsement on Young Vietnamese Consumers‟ Purchasing Intention The Journal of Asian Finance, Economics and Business, 8(1), 951–960 https://doi.org/10.13106/JAFEB.2021.VOL8.NO1.951
Sun, Y., Shao, X., Li, X., Guo, Y., & Nie, K (2019) How live streaming influences purchase intentions in social commerce: An IT affordance perspective
Electronic Commerce Research and Applications, 37, 100886 https://doi.org/10.1016/j.elerap.2019.100886
Wongkitrungrueng, A., & Assarut, N (2020) The role of live streaming in building consumer trust and engagement with social commerce sellers Journal of Business Research, 117, 543–556 https://doi.org/10.1016/j.jbusres.2018.08.032
Xu, X., Wu, J.-H., & Li, Q (2020) WHAT DRIVES CONSUMER SHOPPING BEHAVIOR IN LIVE STREAMING COMMERCE? 21(3), 24
Zhao, N., Zhou, M., Shi, Y., & Zhang, J (2015) Face Attractiveness in Building Trust: Evidence from Measurement of Implicit and Explicit Responses Social Behavior and Personality: An International Journal, 43(5), 855–866 https://doi.org/10.2224/sbp.2015.43.5.855
Zhu, B., Kowatthanakul, S., & Satanasavapak, P (2019) Generation Y consumer online repurchase intention in Bangkok: Based on Stimulus-Organism-Response (SOR) model International Journal of Retail & Distribution Management, 48(1), 53–69 https://doi.org/10.1108/IJRDM-04-2018-0071
Zhu, L., Li, H., Wang, F.-K., He, W., & Tian, Z (n.d.) How online reviews affect purchase intention: A new model based on the stimulus-organism- response (S-O-R) framework 26