Mechanisms of Consumer Purchase Intent Formation in the Context of Live Streaming E-commerce

: This paper investigates the factors influencing purchase intent in the context of live streaming e-commerce and the variations in their impact among consumers with different levels of digital literacy. By constructing a purchase intent model for live streaming e-commerce, we collected 248 valid survey responses through a questionnaire survey. We employed a structural equation modeling approach to verify the influence of five dimensions of perceived value on purchase intent in live streaming e-commerce and the moderating effect of digital literacy. The results indicate that consumers' perceived value significantly affects their purchase intent through their attitudes toward live streaming e-commerce. The order of the impact of perceived value dimensions on consumer attitudes is as follows: entertainment, supply chain perception, functionality, economy, and host characteristics. Notably, as consumers' digital literacy increases, they tend to focus more on the supply chain perception factor.


Introduction
According to commercial big data, in the first half of 2020, there were over 10 million live streaming e-commerce sessions, with more than 400,000 active hosts and over 50 billion viewers.The number of products available for sale exceeded 20 million [1].In just 26 seconds after the start of the 2020 Double 11 (Singles' Day) shopping festival, the peak order creation rate on Tmall reached 583,000 orders per second.By 24:00 on November 11, 2020, the real-time logistics order volume for the Tmall Double 11 Global Shopping Festival was 2.321 billion, with a transaction volume of 498.2 billion RMB, and JD's live streaming sales reached 271.5 billion RMB [2].
From 2016 to the present, live streaming e-commerce has experienced nearly five years of development, demonstrating a strong growth trend.It has become increasingly integrated into people's daily lives.However, what are the mechanisms that attract consumers to live streaming ecommerce?Why is it favored by consumers?The COVID-19 pandemic in 2020 accelerated the development of live streaming e-commerce, but it also brought about various challenges, such as data fraud and the sale of counterfeit goods.In the post-pandemic era, is live streaming e-commerce a passing trend or a sustainable development?To answer these questions, it is necessary to study the development mechanisms of live streaming e-commerce and propose corresponding strategies.

Literature Review
Currently, in academia, research on consumer purchase intent in traditional shopping scenarios primarily examines consumer attitudes [3], maximizing perceived value [4], and minimizing perceived risk [5].It is widely acknowledged that consumer characteristics, product lines, consumption scenarios, and socioeconomic factors are crucial factors influencing consumer purchase intent [6].The characteristics of online products do share similarities with traditional shopping, but the key difference lies in the fact that online products cannot provide consumers with direct sensory stimulation [7].Instead, consumers can only gain insights through static cues like text and images, lacking the tangible experience of touch.
In the study of consumer online purchase intent, in addition to consumer attitudes, product cues, and perceived risk, scholars have expanded their research from the perspectives of customer trust [8], transaction costs [9], social presence [10], among others.Factors affecting consumer online purchase intent mainly include website environment: webpage design [11], technological maturity, and visual effects [12], website ambiance [13], and more.Kim and others suggest that multidimensional display technologies such as 3D can offer consumers comprehensive product information, enhancing the shopping experience and participation, thereby increasing purchase intent [14].Diao Lei Yu believes that website convenience, aesthetics, and the perceived utility of information have a positive impact on consumer purchase intent [15].Furthermore, many researchers have studied the impact mechanisms of online influencers on consumer purchase intent in live streaming e-commerce.Building on this, subsequent research posits that besides influencer factors, product information, promotional elements, personalized services, and perceived value elements all influence consumer purchase intent [16][17][18][19].
Scholars have conducted considerable research on live streaming e-commerce, providing a valuable foundation for this paper.However, live streaming e-commerce is an emerging phenomenon, distinct from the factors affecting consumer purchase intent in the traditional era of text and images.It exhibits characteristics such as a dynamic, visual, and scenario-driven nature, and should be considered holistically, encompassing the frontend (live streams, hosts, product features, pricing, etc.) and backend (supply chain factors like payment, logistics, after-sales, etc.).In the context of the new digital economy, as consumers' digital literacy gradually improves, how does this variable affect their decision-making?These areas are still ripe for exploration.Therefore, this study aims to address the following research gaps: (1) What are the mechanisms by which live streaming e-commerce attracts consumers to generate purchase intent?(2) How does consumer digital literacy moderate the relationship between perceived value and attitude?Thus, this study attempts to answer these questions by constructing a multi-channel retail system adoption intention model.

Hypothesis Formulation
"Economy" refers to consumers finding products of interest to be cheaper than their expectations [20].The economy of a product can reduce the pain associated with payment [21].Existing research indicates that cost is a significant influencing factor in the adoption behavior of information systems, significantly impacting users' attitudes toward adopting services [22].Consumers compare the purchase costs of live streaming e-commerce and use this as a decision-making factor.The economy of a product implies reduced purchase costs.When consumers perceive that shopping through live streaming e-commerce will save them money, they are more likely to make purchases through this channel.Therefore, the following hypothesis can be proposed: H1: Economy has a positive impact on consumer attitudes.
When the sales theme is clear, the direct presentation of commercial information such as brands, products, prices, and promotional methods during live streams is advantageous in saving consumers time in searching and comparing, thus increasing decision efficiency [23].Hosts rely on their professional skills to demonstrate and introduce products to consumers, providing differentiated and specialized services that facilitate comprehensive understanding of product usage and functionality, making it easier for consumers to make purchase decisions.Based on this, the following hypothesis regarding the impact of functionality on consumer attitudes can be proposed: H2: Functionality has a positive impact on consumer attitudes."Entertainment" refers to "the degree of pleasure obtained when perceiving or enjoying a product or service without the influence of other factors" [24].Wexler pointed out that the perception of entertainment significantly influences computer users' acceptance of new systems.Furthermore, Lee's research showed that the perception of entertainment significantly improves customers' attitudes toward retailers.Therefore, when customers discover that shopping through live streaming ecommerce provides a joyful shopping experience, their attitudes toward using live streaming ecommerce become more positive.Hence, the following hypothesis can be proposed: H3: Entertainment has a positive and stimulating impact on consumer attitudes.Opinion leaders have a more significant and lasting impact on consumer purchase intent than traditional marketing, especially in online virtual communities, where opinion leaders can influence consumer behavioral intentions through their expertise, product involvement, interaction, and consistency.Internet celebrity hosts combine the characteristics of celebrities and opinion leaders, having a large fan base and a certain level of influence.In online communities, they act as opinion leaders, influencing community members through professional knowledge, high-quality speech, and frequent interaction, thus having a significant impact on consumers' purchase intent [17].Therefore, the following hypothesis can be proposed: H4: Host characteristics have a positive and stimulating impact on consumer attitudes.The live streaming industry chain consists of upstream brand companies, midstream operating agencies, and downstream content and e-commerce platforms.The information generated through their cooperation forms the content of live streaming.Part of the live streaming content is directly presented in the live room, using forms such as text, graphics, sound, tone, and other forms to present information-type content and emotional content to consumers.Another part exists covertly in transaction links, payment methods, logistics delivery, after-sales services, and other processes, objectively affecting consumers' rational consumption behavior [23].Based on the differences in supply chain service content in the live streaming backend, the following hypothesis can be proposed: H5: Supply chain perception has a positive and stimulating impact on consumer attitudes."Attitude" refers to an individual's enduring evaluation, feeling, and tendency toward certain individuals or concepts, while "intent" reflects the strength of an individual's intention to perform specific actions [24].The Theory of Reasoned Action (TRA), Theory of Planned Behavior (TPB), and Technology Acceptance Model (TAM) all suggest that users' attitudes toward the use of a new system directly influence their intention to use that system.This viewpoint is supported by numerous empirical studies [27].Therefore, it can be assumed that customers' attitudes toward multi-channel retail systems should directly influence their intention to adopt the system.Thus, the following hypothesis can be proposed: H6: Consumer attitudes have a positive impact on purchase intent.The term "digital literacy" first appeared in 1994 when Israeli scholar Y. Eshet-Alkalai summarized it as the "ability to understand and use a variety of digital resources and information displayed through computers" [28].In 2004, he outlined five aspects of digital literacy: visual literacy, innovation literacy, branch literacy, information literacy, and socio-emotional literacy.Subsequently, Paul Gilster formally introduced the concept of "digital literacy" in his work "Digital Literacy," emphasizing a critical evaluation of online information [29].As one of the eight core literacies emphasized by the European Union, digital literacy has evolved from its initial focus on "information technology" (emphasizing the technology itself) to "information technology skills and usage," "information technology skills" (emphasizing usage skills), and finally "digital literacy" (emphasizing literacy) [30].In 2012, the American Library Association (ALA) Digital Literacy Task Force defined digital literacy as the "ability to use information and communication technologies to find, evaluate, create, and communicate digital information, requiring both cognitive and technical skills" [31].
In the digital age, consumer digital literacy mainly pertains to digital information, emphasizing the development of people's ability to use digital tools, understand multimedia digital information, and foster innovative and critical thinking.Live streaming e-commerce, as an emerging phenomenon of the digital age, is received differently by consumers with varying levels of digital literacy.Hence, the following hypothesis can be proposed: H7: Consumer digital literacy moderates attitudes.

Conceptual Model Construction
Purchase intent refers to the probability of consumers purchasing goods or services, encompassing both positive and negative dimensions.Consumer purchase intent plays a critical predictive role in consumer behavior.According to Fishbein & Ajzen's Theory of Reasoned Action (TRA) [31], an individual's beliefs about the outcomes of a behavior, including their evaluation of the value of those outcomes, influence the individual's attitude towards performing that behavior, subsequently affecting their intent to act.Building upon the Theory of Reasoned Action, Moon & Kim introduced and validated the extended Technology Acceptance Model (TAM) under internet contexts [27].In this model, perceived usefulness, perceived ease of use, and perceived enjoyment are three critical beliefs that affect consumer attitudes and intentions to use.Given that the live streaming e-commerce environment is a technology-driven transaction setting, this study posits that these three beliefs can also influence consumer attitudes towards live streaming e-commerce, thereby affecting their purchase intent.These beliefs include functionality, economy, and entertainment.Furthermore, based on scholars' research, the model incorporates the host characteristics as a frontend factor and the "perceived supply chain" as a backend factor.Based on the above analysis, digital literacy is introduced as a moderating variable into the model, and it is believed that this variable can moderate the relationship between consumer perceived value and attitude towards using live streaming e-commerce.Following the impact process of "customer perceived value (beliefs) -usage attitude -adoption intent."

Research Design
Based on the analysis of consumer perceived value in the express delivery industry, the questionnaire was divided into two parts: the first part collected basic information about the respondents, including gender, age, occupation, educational background, and more.The second part consisted of the Consumer Perceived Value Scale, which was divided into five dimensions (functionality, economy, entertainment, host characteristics, and perceived supply chain).Each dimension had 3-4 questions, as shown in Table 1.The second part of the questionnaire formed the fundamental data for the subsequent analysis.Both qualitative and quantitative questions in the questionnaire used a 5-point Likert scale to ensure the uniformity and consistency of the questionnaire.

Q10
The host's ability to promote products is professional.Zhu Tong (2017), Wu Yuhong (2017) Q11 The host is interesting and has personal charm.
Q12 The host has strong interaction skills with fans.

Perceived Supply Chain
Q13 Live streaming shopping allows secure payments.

Batachij et al. Q14
Live streaming shopping provides worry-free after-sales service.

Q15
I can shop on live streams with complete peace of mind.

Q18
Live streaming e-commerce shopping is very appealing to me.

Q20
When I encounter useful digital information, I have the habit of recording it.

Q21
When I need to buy products online, I know what relevant knowledge I should have.

Q22
When I need to purchase products, I can find live streams independently.I am very satisfied with the services provided by the live streaming e-commerce platform I use.

Q25
The platform's services exceed my expectations, and I will continue to choose this live streaming platform in the future.After the initial questionnaire draft was prepared, the project team submitted the questionnaire to relevant experts at the school for review.Adjustments were made to address vague and inaccurate phrasing and to make certain questions more concise, ensuring greater clarity and reducing the potential for ambiguity.
To ensure the credibility of the questionnaire results, the project team conducted a preliminary test questionnaire.The test questionnaire included 24 questions that assessed the sufficiency of each dimension.In terms of distribution, given the ongoing need for contact reduction due to the normalized pandemic control measures, this study used an online method to distribute the questionnaire.The electronic questionnaire was designed to prevent multiple entries and ensure data quality.The electronic questionnaire was distributed through social networks to students, friends, and others, ensuring the timeliness of questionnaire sources.In the end, a total of 302 experimental questionnaires were distributed, with 249 responses collected, resulting in an effective questionnaire rate of 82.4%..74%From the data in the table, it is evident that the majority of respondents were female, accounting for 61.5% of the sample, indicating a higher level of female participation in live streaming ecommerce.Respondents were primarily between the ages of 18 and 40, representing nearly 97% of the total sample.Among them, the age group of 19-25 years accounted for 63%, suggesting that this age group is the main user base for live streaming e-commerce.Geographically, over 80% of the respondents lived in tier 1 cities.In terms of education, most respondents had at least a university degree, with a significant portion holding postgraduate degrees.

Reliability Test
In this study, as all the variables in the measurement scale were assessed using a 5-point Likert scale, the reliability of the scale was evaluated using Cronbach's α coefficient.It is generally accepted that a higher Cronbach's α coefficient indicates greater internal consistency of the scale.An α value greater than 0.7 suggests relatively high reliability, while values between 0.35 and 0.7 indicate acceptable reliability.Values below 0.35 suggest low reliability and should be rejected.From Table 3, it can be observed that all variables have Cronbach's α coefficients above 0.8, indicating that the questionnaire has good reliability.Furthermore, a Kaiser-Meyer-Olkin (KMO) test and Bartlett's test of sphericity were performed on the questionnaire.The KMO test yielded a coefficient of 0.937, exceeding the recommended threshold of 0.5, and the statistical significance probability (p-value) was less than 0.001, which is lower than the critical p-value of 0.05.This indicates that the structural validity test was passed, allowing for factor analysis to be conducted.889 In this study, structural equation modeling was conducted using AMOS software to verify the relationships between these variables.Based on the hypotheses of this study and the conceptual model presented, a complete structural equation model to be tested was established using AMOS.

Confirmatory Factor Analysis
Confirmatory factor analysis was conducted using AMOS 24.0 and SPSS 26.0 software to validate the factors.The software calculations revealed that the minimum factor loading between the measured variables and latent variables was 0.728, which is greater than 0.6.Therefore, no items needed to be deleted.The composite reliability (CR) and average variance extracted (AVE) for each factor were calculated.The CR values for Functionality, Economic Value, Entertainment, Host Traits, Supply Chain Perception, Consumer Attitude, and Purchase Intention were found to be 0.618, 0.697, 0.813, 0.696, 0.676, 0.599, and 0.626, respectively, all of which exceeded 0.5.The AVE values for these factors were 0.829, 0.872, 0.929, 0.872, 0.86, 0.817, and 0.834, respectively, all exceeding 0.8.This indicates that each factor has good convergent validity, as shown in Table 4.  From Table 8, it can be observed that when consumers have lower digital literacy, they are more concerned with the functionality and entertainment value of products in live e-commerce.However, when consumers have higher digital literacy, they pay more attention to the impact of supply chain factors.Additionally, consumers' focus on functionality and entertainment decreases relatively.These findings support H7.

Conclusion
This article explores consumers' purchase intention in the context of live e-commerce from the perspective of perceived value, while also verifying the moderating effect of consumer digital literacy.The following research conclusions are drawn: 1. "Economic value," "entertainment," "functionality," "host traits," and "supply chain perception" significantly influence consumers' attitudes toward live e-commerce."Functionality" indicates that meeting basic usability needs is a crucial motivator for consumer purchase intention.Consumers are more inclined to favor products that are not only functional but also economically advantageous and cost-effective.Furthermore, due to the diverse nature of consumer shopping motivations, in addition to functionality and economic value, more emphasis should be placed on the entertainment factor brought by live e-commerce.Leveraging host traits to enhance consumer attraction is essential.Attention should be given to improving consumers' perception of the supply chain.In other words, a comprehensive understanding of consumer purchase motivations is needed to provide a holistic live e-commerce shopping experience.
2. Consumers' positive attitudes toward live e-commerce significantly affect their purchase intention.This conclusion aligns with the findings of previous scholars and emphasizes that consumer usage attitudes are decisive factors in the adoption of a multi-channel retail system.[27] Therefore, enhancing consumers' perceptions of value to improve their attitudes toward a multi-channel retail system should be a crucial approach for retailers to strengthen consumer adoption intentions.
3. Consumer digital literacy significantly moderates the relationship between perceived value and attitude.For consumers with lower digital literacy, the "entertainment" value of live e-commerce and the functionality of displayed products are the primary factors affecting their attitude.Among these factors, "entertainment" holds the highest weight.In other words, when consumers have lower digital literacy, enhancing the entertainment value of live e-commerce can easily improve their attitudes.For consumers with higher digital literacy, "supply chain perception" has the strongest impact on their attitude toward live e-commerce.Providing higher-quality supply chain services can positively change the attitude of this group toward live e-commerce.

Research Limitations and Prospects
(1) The factors explored in this paper that affect purchase intention in live e-commerce primarily focus on perceived value.Further research could introduce other variables such as "trust" into the model to enhance its explanatory power.
(2) This study only examined the moderating effect of digital literacy.Subsequent research could consider other factors like price sensitivity as moderating variables to understand more factors that affect purchase intention due to differences in consumer characteristics.
(3) This research solely employed a questionnaire survey to gather data.Further studies could design more rigorous experimental procedures to validate the effectiveness of the model proposed in this research.
streaming e-commerce shopping is a clever way to shop.

Table 1 :
Measurement Dimensions and Reference Sources

Table 2 :
Summary of Consumer Demographic Information

Table 3 :
Descriptive Statistics and Reliability Test of the Sample Scale

Table 4 :
Results of Confirmatory Factor Analysis

Table 6 :
Moderation Effects Analysis