The Sustainability of Fresh Agricultural Produce Live Broadcast Development: Influence on Consumer Purchase Intentions Based on Live Broadcast Characteristics

Abstract

This paper is based on consumer purchase decisions and uses the consumers of live fresh agricultural products as the research object. We analyze consumers’ intention to purchase and analyze the role of perceived risk and value co-creation in consumers’ purchase of fresh agriculture products. Structural equation modeling analysis and stepwise regression coefficient analysis were used. The results show that: (1) the fresh agricultural products’ live features can positively influence consumers’ intention to buy; (2) the authenticity of fresh agricultural products live broadcast shows a significant negative correlation with perceived risk, and the visibility of fresh agricultural products live broadcast shows a significant positive correlation with value co-creation; (3) the perceived risk and value co-creation mediate between the live-stream features of fresh agricultural products and consumers’ intention to buy. The study results provide a basis for management decisions on the live broadcast of fresh agricultural products. It also has important implications for the sustainability of live broadcast of fresh agricultural products.

1. Introduction

The growth of China’s Internet industry has accelerated the development of live broadcast [1]. Everyone is gradually discovering the value of live broadcast, and fresh agricultural produce merchants join in live broadcast [2]. There is a big difference between traditional e-commerce shopping for fresh agricultural produce and live shopping. Traditional e-commerce shopping showcases product images, while live shopping combines multi-dimensional interaction and real-time video display. The change in selling format enhances the customer’s shopping experience and changes the consumer’s willingness to buy [3]. Fresh agricultural products live broadcast is an online shopping method. Farmers or professional web hosts act as the anchors of the live broadcast, and they are responsible for promoting and selling agricultural products. Consumers enter each fresh agricultural product’s live broadcast room through the e-commerce platform to make purchases. Fresh agricultural products are easily damaged and perishable, making fresh agricultural live broadcast a particular type of live broadcast in live marketing [4]. Therefore, it is essential to clarify the characteristics of fresh agricultural products live broadcast, analyze consumers’ intention to purchase fresh agricultural products live broadcast as well as the relationship between them for the sustainable development of fresh agricultural products live broadcast.

Fresh agricultural products include fresh vegetables, fruits, meat, aquatic products, eggs and milk, live poultry, and primary processing products [5]. Because of its perishable product characteristics, consumers risk prejudging the live purchase of fresh agricultural produce. Douglas et al. (2012) argue that merchants need an actionable framework to guide consumers toward value co-creation initiatives [6]. Achieving value exchange between consumers and merchants increases loyalty to the live stream [7]. Perceived risk and value co-creation have proven to be particularly important in the impact of live fresh agricultural products on consumer purchase intentions.

This paper analyzes the fresh agricultural products live broadcast characteristics and constructs the fresh agricultural products live broadcast characteristics scale comprehensively and systematically, which will enrich the research of fresh e-commerce live broadcast characteristics to a certain extent, and comprehensively grasp the influencing factors of fresh agricultural products live broadcast characteristics on consumers’ purchase intention through the fresh agricultural products live broadcast characteristics scale. Moreover, the research in this paper also expands the application of the SOR theoretical model in the field of live e-commerce and introduce perceived risk and value co-creation into the analysis of the factors influencing purchase intention of fresh agricultural products live, to enrich the influencing factors of live e-commerce characteristics. Lastly, the research proposes operational and management suggestions for living fresh agricultural produce merchants to promote the sustainable development of live fresh agricultural products in China.

2. Literature Review

2.1. A Research on Live Broadcast of Fresh Agricultural Products

The current research directions of fresh agricultural products live broadcast are generally divided into three categories: research on the supply chain model, development model, and consumer purchase intention.

In the fresh agricultural products live supply chain model study, scholars have increased the profitability of online agricultural products sold by updating the supply chain model and optimizing logistics technology [8] or using cluster-based road heuristics (CRH) to study the transport routes of fresh logistics and optimize the supply chain [9]. Rong (2011) proposed a linear programming mathematical model with mixed integers to control the current fresh agricultural supply [10]. A study on the development of a live model for fresh agricultural produce found that live broadcast on the Internet is entertaining, interactive, experiential, and inducing [11]. The story of fresh agricultural products has some challenges, such as the severe phenomenon of “homogenization,” inadequate laws and regulations, and the quality management of fresh food. These issues have become the focus of scholars’ research [12]. In the study of purchase intentions for fresh agricultural products live-broadcast, scholars found that only 56.4% of consumers in the live-broadcast room said they were willing to spend [13]. Improving consumers’ purchase intention in fresh agricultural products live broadcast is essential in promoting the sustainable development of fresh agricultural products live broadcast.

At present, the research on consumers’ willingness to purchase is mainly focused on consumer characteristics, and there is a lack of research on the sustainability of fresh product development with the fresh elements of fresh agricultural produce e-commerce as the entry point.

2.2. Research on The Purchase Intention of Fresh Agricultural Products Live Consumers

Wu, C (2014) believes that in consumer e-commerce, factors such as age, income, education level, leisure time sufficiency, and consumers’ online purchasing experience all play an essential role in influencing their purchasing behavior [14]. The relationship between consumers’ psychological and behavioral traits affects consumers’ judgments about purchase intentions [15]. Analysis of consumer characteristics will facilitate the sale of fresh agricultural products. Another group of scholars considers the influence of online reviews [16], website shopping orientation [17], and Internet word-of-mouth [18] on consumers’ intention to purchase agricultural products on e-commerce platforms. If the e-commerce website contains informative, easy-to-understand product information, it can positively impact consumers’ choice to buy products online [19]. The live broadcast with goods model is the reason why consumers are willing to buy goods [20]. Some other scholars believe that the quality of fresh agricultural products and the delivery time of fresh logistics are the main concerns of consumers and the most important reason for consumers to decide whether to buy [21]. Although fresh agricultural products e-commerce has won the hearts of consumers, the quality of fresh agricultural products is still the primary criterion for consumers to buy [22].

In combing the literature, scholars found that the research on consumer purchase intention of fresh agricultural products live broadcast was conducted from three aspects: consumer characteristics, e-commerce live broadcast platform characteristics, and product attribute characteristics. There is a lack of research on live-broadcast fresh agricultural products affecting consumers’ purchase intention.

2.3. Research Research on Perceived Risk and Value Co-creation on Fresh Agricultural Products Live Broadcast Consumers’ Purchase Intentions

The fresh agricultural products live broadcast has increased because of the novelty of the sales method and the specificity, and rational consumers will collect product information and reduce risks when buying online [23]. The attributes of the product shown by the host in the live broadcast, the consumer’s familiarity with purchasing online goods, and the consumer’s perception of the product will all affect the perceived risk [24]. Perceived risk is divided into financial, product, delivery, service, and psychological aspects. Perceived economic, development, and service risk significantly affect customers’ intention to shop online, while distribution and psychological risks have no considerable impact [25]. Liang Yuan et al. (2017) analyzed the perceived risks in six aspects, including physical, behavioral, financial, temporal, social, and psychological. It was concluded that perceived risks could cause consumer anxiety and negative attitudes toward online shopping [26].

Fresh agricultural products live merchants and consumers describe product information or opinions to form an information exchange. In the live broadcast of fresh agricultural products in value co-creation, merchants are no longer aspects of the release of information but two-way communication with consumers in the live broadcast. Consumers with ask questions make merchants understand the needs of consumers and better serve them [27]. Payne (2008) suggests that self-directed consumer learning positively affects company and brand awareness and that this sense of self-directed learning influences consumers’ willingness to buy [28].

Live broadcast of fresh agricultural products has broadened the way consumers buy, but the perceived risk of consumer perception is increased by the inability to inspect fresh goods visually. Therefore, it is essential to reduce the perceived risk of the fresh agricultural produce live broadcast purchase process. Few studies have incorporated value co-creation into the shopping decision mechanism of live fresh agricultural products consumers, and value co-creation should be given more attention as a communication channel to enhance merchants and consumers.

2.4. Research Review

From the existing studies, few scholars have studied consumer purchase intentions using the live characteristics of fresh agricultural products as an entry point. Perceived risk and value co-creation have not been jointly included in the study of fresh agricultural products live consumer shopping decision mechanisms. In summary, this paper will create a scale of fresh agricultural live broadcast characteristics from the perspective of fresh agricultural live broadcast sustainability and analyze the influence of fresh agricultural live broadcast characteristics on consumers’ purchase intention.

3. Data and Methods

3.1. Theory Model

Consumer purchase decision means that consumers will choose a product only after judging the information about the product. The primary consumer purchase decisions are the stimulus–organism–response (SOR) model, the Kotler behavioral choice model, the Nicosia model, the Engel model, and the Howard–Sheath model. The SOR model has been widely used in the study of consumer purchase intention. Nadjim M (2021) used the SOR model to study the consumer purchase intention of dairy products. The effects of perceived quality, price, and satisfaction on purchase intention were explored [29]. Peng Z (2019) explored the mechanism of purchase intention influence when purchasing travel products online based on the SOR model, analyzing the effect of perceived risk on purchase intention [30]. Therefore, the SOR model is also selected in this paper as a model to explore the mechanisms of live broadcast and consumer purchase intention for fresh agricultural products. The perceived risk theory suggests that any purchase behavior of consumers may not know for sure whether their expected results are correct or not, and some results may be unpleasant to consumers. Therefore, this paper will verify the mechanism of the role of perceived risk between the live characteristics of fresh produce and consumers’ willingness to buy. The social exchange theory is based on the premise that all people are “economic people” and believes that the behavior of all people is controlled by some kind of exchange activity, which can bring explicit or implicit returns. When the consumers in the fresh produce live broadcast booth share information with other consumers, exchange emotions, and gain friendship and a sense of belonging, they will perceive the value brought by the fresh produce live broadcast booth when they interact with other consumers and obtain relevant resources. For reciprocity, consumers will be willing to help other consumers or participate in activities that are beneficial to businesses. Consumer behavior determines the outcome of value co-creation between consumers and companies, and value co-creation affects purchase intention. This paper verifies the mechanism of value co-creation between the characteristics of fresh produce live broadcast and consumers’ willingness to purchase. In the live broadcast of fresh agricultural products, consumers cannot immediately perceive the quality of agricultural products, which makes consumers lack trust, because it is different from consumers’ previous purchase of agricultural products in the field. The interactive form of fresh agricultural products live broadcast gives consumers and merchants more opportunities and space for value co-creation.

In this paper, the SOR model fully captures the activity of consumer purchase intentions in live fresh agricultural products. S is the external stimulus (the stimulus of watching fresh agricultural products live); O is the organism, which contains the consumer’s consciousness (perceived risk) and emotional variables (value co-creation); and R is the reaction (the psychological change of the consumer after being stimulated by the external world). This study aims to expand the value co-creation and perceived risk mechanism based on the SOR model. Discovering its role in the fresh agricultural produce live consumer purchase intention model, the underlying theoretical model of this study is shown in Figure 1.

3.2. Research Hypothesis

3.2.1. The Effect of Live Characteristics of Fresh Agricultural Products

The live broadcast of fresh agricultural products is a large category of many webcasting classifications, so the live broadcast characteristics are similar to those previously studied by scholars. H.Ma (2018) explores online live broadcast and finds that live broadcast of fresh agricultural produce is visual, interactive, authentic, and entertaining [31]. Real-time visibility of fresh agricultural products includes the visual exposure of the consumer to the products while communicating with the anchor. We use text, sound, images, and other forms to convey product information based on online video communication. We stimulate customers’ experience through multiple senses and improve their immersion. Fifth generation mobile communication technology (5G) technology in the media industry has led to more precise and realistic live picture quality. According to the SOR theory, when consumers are attracted to the current product, they will ignore other information, thus increasing the customer’s perceived value. Most live broadcasts of fresh agricultural products are conducted in the planting area. Through the anchor’s live experience and display, consumers can experience the actual state of fresh agricultural products. Live broadcast is a buy-while-you-watch marketing method that gives consumers a feeling similar to a physical store. Consumers can see the product’s real value in the anchor’s trial and evaluation, thus improving the consumer’s impression of the product [32]. Hypotheses are formulated based on the above analysis.

The impact of the fresh agricultural live characteristics on consumer perceived risk is as described above. Because of the change of sales method, consumers will perceive the risk involved. The anchor’s product introduction, Q&A interaction, etc. will change the consumer’s perceived risk so that the consumer has a significant positive effect on perceived finance, product, delivery, and service.

Hypothesis H1.

The live characteristics of fresh agricultural products have a significant adverse effect on perceived risk.

Hypothesis H1a.

Live broadcast visibility of fresh agricultural products significantly negatively affect perceived risk.

Hypothesis H1b.

Live broadcast interactivity of fresh agricultural products significantly negatively affect perceived risk.

Hypothesis H1c.

Live broadcast authenticity of fresh agricultural products significantly negatively affect perceived risk.

Hypothesis H1d.

Live broadcast entertainment of fresh agricultural products significantly negatively affect perceived risk.

Hypothesis H2.

Live broadcast characteristics of fresh agricultural products have a significant positive impact on value co-creation.

Hypothesis H2a.

Live broadcast visibility of fresh agricultural products has a significant favorable influence on value co-creation.

Hypothesis H2b.

Live broadcast interactivity of fresh agricultural products has a significant favorable influence on value co-creation.

Hypothesis H2c.

Live broadcast authenticity of fresh agricultural products has a significant favorable influence on value co-creation.

Hypothesis H2d.

Live broadcast entertainment of fresh agricultural products has a significant favorable influence on value co-creation.

Hypothesis H3.

Live broadcast characteristics of fresh agricultural products have a significant positive effect on purchase intention.

Hypothesis H3a.

Live broadcast visibility of fresh agricultural produces significantly positively affects purchase intention.

Hypothesis H3b.

Live broadcast interactivity of fresh agricultural produces significantly positively affects purchase intention.

Hypothesis H3c.

Live broadcast authenticity of fresh agricultural produces significantly positively affects purchase intention.

Hypothesis H3d.

Live broadcast entertainment of fresh agricultural produces significantly positively affects purchase intention.

3.2.2. Influential Effects of Perceived Risk

Perceived risk refers to the negative outcome of consumers due to their lack of confidence in the goods or services. In addition, some researchers have argued that perceived risk refers to the uncertainty and the resulting outcomes when consumers are faced with shopping decisions. Consumers have a higher perceived risk when they pay more attention to this negative information, they have a higher perceived risk [33]. Scholars generally agree that perceived danger is not a one-way street but is composed of the following factors. The financial crisis is a danger related to physical or private goods, mainly reflected in the perception of insecurity of payment behavior. The behavioral risk is the potential for insecurity when the function of the goods or services is satisfied. This situation is strongly associated with the brand of the e-commerce platform, as the more reputable platform can endorse the quality of the product or service, which in turn enhances consumer trust and reduces consumer perceived risk. Physical hazards are the awareness of hazards related to health or energy loss [34]. Hypotheses are formulated based on the above analysis.

Hypothesis H4.

Perceived risk has a significant negative effect on purchase intention.

Hypothesis H5.

Perceived risk plays a mediating role between fresh agricultural produce live broadcast characteristics and purchase intention.

Hypothesis H5a.

Perceived risk plays a mediating role between live broadcast visibility of fresh agricultural produce and purchase intention.

Hypothesis H5b.

Perceived risk plays a mediating role between live broadcast interactivity of fresh agricultural produce and purchase intention.

Hypothesis H5c.

Perceived risk plays a mediating role between live broadcast authenticity of fresh agricultural produce and purchase intention.

Hypothesis H5d.

Perceived risk plays a mediating role between live broadcast entertainment of fresh agricultural produce and purchase intention.

3.2.3. Influence Effect of Value Co-creation

Payne et al. (2008) suggest that, as customers become more cognizant of brands and goods, consumers will further understand the companies they identify with and become more interested in the products [28]. The consumer’s evaluation of the product or the advice provided are important factors to measure whether the consumer agrees with the company and can have purchasing behavior. Hypotheses are formulated based on the above analysis.

Hypothesis H6.

Value co-creation has a significant positive effect on purchase intention.

Hypothesis H7.

Value co-creation plays a mediating role between fresh agricultural produce live broadcast characteristics and purchase intention.

Hypothesis H7a.

Value co-creation plays a mediating role between live broadcast visibility of fresh agricultural produce and purchase intention.

Hypothesis H7b.

Value co-creation plays a mediating role between live broadcast interactivity of fresh agricultural produce and purchase intention.

Hypothesis H7c.

Value co-creation plays a mediating role between live broadcast authenticity of fresh agricultural produce and purchase intention.

Hypothesis H7d.

Value co-creation plays a mediating role between live broadcast entertainment of fresh agricultural produce and purchase intention.

3.3. Scale Design

3.3.1. Scale Design for Live Characteristics of Fresh Agricultural Products

According to the analysis of the characteristics of live broadcast of fresh agricultural products in the literature review, the parts of live broadcast of fresh agricultural products are broadly divided into interactivity, authenticity, visibility, and entertainment. Previous studies have widely used the characteristics of e-commerce live broadcast to analyze fresh agricultural produce live broadcast, and it remains to be explored whether this approach can reflect its characteristics more comprehensively and accurately. Fresh agricultural products are generally perishable, and effective, and purchase mainly depends on consumers’ purchasing experience. Because it is different from ordinary live e-commerce products, consumers need to present additional requirements for anchors. Using the generic e-commerce live broadcast scale is no longer sufficient to fully explain the phenomena in fresh agricultural products live broadcast. To sum up, this paper believes that we will develop a hierarchy of live broadcast characteristics of fresh agricultural produce and analyze the influence of live broadcast characteristics of fresh agricultural produce on consumers’ purchase intention based on the scale. It will make this study more rigorous and scientific.

To obtain the characteristics of fresh agricultural products live broadcast and ensure the sample’s representativeness, this study focuses on consumers who have purchased fresh agricultural products in fresh agricultural products live broadcast as the interviewees. A total of 30 consumers were interviewed. Yin R (2009) pointed out that the conclusions reached by different channels and subjects remained similar, indicating the credibility of the data [35]. The types of interviewees in this study are divided into two parts. The first part is to use school students as interviewees, and the second part is to select some representative interviewees. Considering that fresh agricultural produce live consumers have the characteristics of balanced distribution in terms of region, industry, and education level, most of their ages are concentrated between 20 and 50 years old. Therefore, the conditions of the interviewees are controlled in this aspect. The interview form was mainly chosen to be conducted in telephone interviews. In terms of the interview process, the entire interview process was recorded after the interviewees’ consent was sought.

A panel of six experts, consisting of two professors, two associate professors, and two PhDs, evaluated the extracted question items for readability and match of topic types obtained in the second phase. Based on the opinions of the professionals, the original questions were modified, and the final questionnaire on the live characteristics of fresh agricultural products was formed.

This study was conducted in June 2021, using the Questionnaire Star software to distribute questionnaires online to consumers who had purchased fresh agricultural produce in the fresh agricultural produce live broadcast. A total of 123 questionnaires were collected for this research on the characteristics of live broadcasts of fresh agricultural produce. After manual censoring by the author, 120 valid questionnaires were finally obtained, with an efficiency rate of 97.5%. The data analysis process was borrowed from Anselm S (2006) [36]. The method was divided into three steps: open coding, spindle coding, and selective coding. After completing the scale development and questionnaire testing, the scale question wording was slightly modified according to the problems reflected in the survey process and results. Finally, it formed a scale for measuring the live characteristics of fresh agricultural products, as shown in

Table 1. Measurement scale of live characteristics of fresh agricultural product.

Variables	Title Number	Title Content
F1 Visual	VA1	The live broadcast is visually pleasing
	VA2	The live broadcast room shows a visually pleasing design
	VA3	The layout of the live broadcast interface is appealing
F2 Interactive	IA1	I can interact with the anchor in the live broadcast to enhance my understanding of the product
	IA2	I can interact with other consumers in the live broadcast to get feedback on the product from other consumers
	IA3	I would like to engage in interaction during the live broadcast
F3 Authentic	AA1	The anchor showed this product from many angles
	AA2	The anchor promoted my more profound understanding of the product during the product’s introduction
	AA3	The anchor’s direct experience with the product deepened my knowledge of the product
F4 Recreational	RA1	Watching the live broadcast of the purchase made me feel interesting
	RA2	I feel relaxed by watching the live broadcast of the purchase
	RA3	Watching a live broadcast of purchase brings me fun

.

3.3.2. Scale Design for Value Co-creation

Value co-creation is when consumers and merchants of fresh agricultural produce live act as the main body of value creation together. The communication between merchants and consumers is very close, and there will be feedback from users in promoting goods. The consumer’s viewpoints complement and cooperate in the design, development, production, and consumption of goods or services to enhance brand and consumer recognition and co-creation of value.

The value co-creation scale was mainly based on the study of Koh and Kim (2004). The level of knowledge sharing in virtual communities is positively proportional to community outcomes, and the loyalty of virtual community service providers to virtual communities is significantly higher [37]. Ridings (2010) understands that the appeal of virtual communities is critical for organizations that want to tap into their vast information potential [38]. Zwass (2010) believes that consumer co-creation of value has become a significant force in the marketplace thanks to the Internet. Autonomous co-creation is an extensive consumer activity that corresponds to the production of value by consumers [39]. As a result, individuals and communities have become a vital and growing productive force in e-commerce. This can be seen in

Table 2. The scale design of value co-creation.

Variables	Title Number	Title Content	Source
Value Co-creation	VC1	I often share my experience with others in the fresh farming live broadcast	Koh & Kim [35], Ridings [36], Zwass [37]
	VC2	I usually start topics on the live broadcast
	VC3	I often actively respond to questions from other consumers on fresh agricultural products live broadcast
	VC4	I often participate in new product idea solicitation, or evaluation activities initiated by the merchants of fresh agricultural products live broadcast
	VC5	I am willing to participate in the interaction during the fresh agricultural products live broadcast

.

3.3.3. Scale Design for Perceived Risk

Fresh agricultural products live broadcast has changed the way consumers traditionally buy fresh agricultural produce. The inability of consumers to visually inspect the merchandise reduces the level of trust and increases the probability of bad results.

The perceived risk scale refers to the consumer involvement profile scale (CIP) scale proposed by Laurent and Kapferer (1985). He argues that the difference in risk outcomes judged by consumers when engaging in consumption leads to different effects of consumption behavior [40]. Roselius (1971) believes that perceived risk refers to the fact that one’s product is not recognized by others and can even be stigmatized. Whether it is a difference in interest or an error in one’s judgment, it can hurt one’s behavior [41]. In this paper, the cultural differences and the characteristics of live broadcasts of fresh agricultural products are appropriately modified according to the former scale to make them easy to understand and answer with four questions. This can be seen in

Table 3. The scale design of perceived risk.

Variables	Title Number	Title Content	Source
Perceived Risk	PR1	I’ll take higher risks when buying fresh agricultural produce in the air	Laurent & Kapferer [40], Roselius [41]
	PR2	Buying fresh agricultural products on live broadcast may be a wrong decision
	PR3	I would be upset if I had a bad buying experience in a live broadcast of fresh agricultural products
	PR4	Choosing the best fresh agricultural produce live broadcast room is a tricky thing to do

.

3.3.4. Scale Design of Consumer Purchase Intention

The scale of consumer purchase intention was referred to in the rankings of Bansa and Voyer (2000). The main focus of the scale is related to three aspects: the significant impact of willingness, the effect of the choice of purchase intention, and the direction of future selection [42]. In this paper, we have made localized adaptations based on the characteristics of live broadcasts of fresh agricultural products. The specific design of the questions can be seen in

Table 4. The scale design of consumer purchase intention.

Variables	Title Number	Title Content	Source
Consumer purchase intention	CPI1	Fresh agricultural products live broadcast helps me a lot in my purchasing behavior	Bansa & Voyer [42]
	CPI2	Fresh agricultural products live broadcast has a significant influence on my decision making
	CPI3	I am willing to discover my favorite fresh agricultural products by watching live broadcasts
	CPI4	Watching fresh agricultural products live broadcasts can stimulate my desire to buy

.

3.4. Questionnaire Design

The respondents of this study were consumers who had purchased fresh agricultural products from the fresh agricultural live broadcast. The questionnaire consisted of three main parts. In the first part, the theme and purpose of this research are introduced to the research subjects. In the second part, in addition to the primary personal information survey, the frequency of purchasing fresh food online during the live broadcast was added as a filtering condition. Some consumers who had not purchased fresh food in the past year were removed from the questionnaire. The third part is the main body of the questionnaire, and this paper uses a 5-point Likert scale for attitude measurement.

3.4.1. Descriptive Statistical Analysis of the Sample

This study questionnaire survey is for consumers who watch fresh agricultural produce live broadcast; not all consumers who care about e-commerce live broadcast can be included in this study survey topic. The survey data is more accurate. The survey form was made, distributed, and collected through the Star Questionnaire platform to create questionnaires, in the WeChat circle of friends to distribute questionnaires, and finally online collection of questionnaires. Finally, after removing invalid questionnaires, 305 valid questionnaires were obtained. Descriptive statistics were analyzed on the sample data. The descriptive statistics mainly included monthly income, education level, occupation, frequency of fresh food purchases in the past year in the live broadcast, and the essential information at the beginning of the question items.

Among the 305 valid questionnaires, the number of men and women was about the same. According to the age structure, the age of respondents is mainly 21–50 years old, accounting for 69.2%, and the most significant proportion of people aged 21–30 and 41–50 years old is 24.6% and 24.3%, respectively. The 21–30-year-olds have just entered the workforce and started to live alone, have some economic power, and are more accustomed to online shopping. Most of the 41–50 respondents’ expenses came from family members. The respondents’ education level is generally high. Most of them have a bachelor’s degree at 48.5%, a college graduation rate of 31.8%, and a master’s degree or above at the lowest rate of 9.2%. In terms of job composition, respondents have a more balanced job distribution. According to the amount freely available each month, the most significant number of respondents, about 31.1%, earned USD 299–747, which is also related to their jobs. Government agencies and professionals all have stable jobs with stable financial resources. College students, however, have a relatively small amount of monthly freely available living expenses. The lowest is the monthly discretionary spending, which is only 17%. In the survey, the customers who shopped at fresh agricultural produce live one to five times a year comprised 38.4%, compared to 61.6% of customers who shopped six or more times. More than half of the consumers regularly shop live for fresh products. This shows that buying fresh food on live broadcast has become a new way of consumption.

This paper compares the data analyzed by descriptive statistics with the “China Fresh Food E-Commerce Industry Research Report 2021” [43], published by Ariadne Consulting. This paper’s data on consumer characteristics were essentially the same as the reported user profiles. The overall fresh food online shopping users are significantly younger, with consumers aged 25–38 accounting for nearly half of all fresh food e-commerce users. In summary, this research sample is more reasonable.

3.4.2. Reliability and Validity Test

The Statistical Product Service Solutions (SPSS) software was used to perform Kaiser–Meyer–Olkin (KMO) and Bartlett’s sphericity tests on the scales to determine whether they were suitable for factor analysis. The test results are shown in

Table 5. Scale reliability and validity tests.

Latent Variables	Observed Variables	Unstd.	S.E.	T-Value	Std FC	p	SMC	CR	AVE
Visual	VA1	1.085	0.067	16.3	0.817	***	0.667	0.89	0.73
	VA2	1.192	0.065	18.433	0.918	***	0.843
	VA3	1			0.824		0.679
Interactive	IA1	1.062	0.064	16.516	0.861	***	0.741	0.882	0.714
	IA2	1.14	0.069	16.538	0.862	***	0.743
	IA3	1			0.811		0.658
Authentic	AA1	0.819	0.04	20.683	0.849	***	0.721	0.909	0.77
	AA2	0.898	0.043	21.066	0.857	***	0.734
	AA3	1			0.925		0.856
Recreational	RA1	0.931	0.059	15.826	0.829	***	0.687	0.873	0.696
	RA2	1.012	0.062	16.279	0.84	***	0.706
	RA3	1			0.833		0.694
Value Co-creation	VC1	0.79	0.055	14.27	0.72	***	0.518	0.91	0.671
	VC2	1.009	0.054	18.642	0.854	***	0.729
	VC3	1			0.846		0.716
	VC4	1.002	0.055	18.215	0.854	***	0.729
	VC5	0.926	0.054	17.178	0.814	***	0.663
Perceived Risk	PR1	1.065	0.068	15.672	0.829	***	0.687	0.893	0.676
	PR2	1			0.804		0.646
	PR3	1.053	0.066	16.012	0.828	***	0.686
	PR4	1.052	0.066	15.994	0.828	***	0.686
Consumer purchase intention	CPI1	0.989	0.045	21.777	0.895	***	0.801	0.926	0.757
	CPI2	1			0.873		0.762
	CPI3	1.006	0.046	21.821	0.895	***	0.801
	CPI4	0.925	0.051	18.233	0.815	***	0.664

Note: Unstd is the unstandardized regression coefficient; S.E. is the standard error; T-Value is the t-test value; Std FC is the factor loading; *** indicates significance at the 0.001 level; SMC is the factor reliability; C.R. is the combined reliability; AVE is the average variance analysis.

. The KMO value is 0.899, more significant than 0.5. The correlation is high and passes Bartlett’s spherical significance test, suitable for factor analysis.

The standardized factor loadings of the observed variables were analyzed using Analysis of Moment Structure (AMOS) to calculate the combined reliability and convergent validity, and the results are shown in Table 5. The standardized factor loadings for all question items exceeded 0.6 and were significant at the 0.001 level. The standardized factor loadings for all question items exceeded 0.6 and were significant at the 0.001 level. The mean-variance precipitations were more critical than 0.5 by the AVE thresholds suggested by Bagozzi R P et al. (1981) [44]. The combined reliabilities were all greater than 0.7, in line with the recommendations of Hair et al. (1998) regarding C.R. thresholds [45]. The scale has good convergent validity and combined reliability.

4. Results

4.1. Correlation Analysis among Variables

The validated factor analysis in the AMOS software analyzed the correlation coefficients and significance. The correlation coefficients between latent variables in this study are shown in

Table 6. Correlation coefficients and the square root of AVE for latent variables.

	Visual	Interactive	Authentic	Recreational	Perceived Risk	Consumer Purchase Intention	Value Co-creation
Visual	0.854
Interactive	0.307 ***	0.845
Authentic	0.405 ***	0.351 ***	0.877
Recreational	0.399 ***	0.468 ***	0.262 ***	0.834
Perceived Risk	−0.493 ***	−0.451 ***	−0.467 ***	−0.464 ***	0.822
Consumer Purchase Intention	0.510 ***	0.499 ***	0.522 ***	0.503 ***	−0.549 ***	0.87
Value Co-creation	0.383 ***	0.353 ***	0.367 ***	0.359 ***	−0.251 ***	0.534 ***	0.819

Note: *** indicates a 0.01 level (two-tailed), significant correlation, and the diagonal value is the square root of AVE.

.

According to the results of the correlation analysis, none of the potential variables correlated more than 0.8, and the possibility of a co-linear problem could be excluded. The results showed a significant negative correlation between visibility, interactivity, authenticity, entertainment, and perceived risk, supporting hypotheses H1, H1a, H1b, H1c, and H1d. Visibility, interactivity, realism, and entertainment are significantly and positively correlated with value co-creation, supporting hypotheses H2, H2a, H2b, H2c, and H2d. There is a significant positive correlation between viewability, interactivity, authenticity, and entertainment and purchase intention, supporting hypotheses H3, H3a, H3b, H3c, and H3d. Perceived risk has a significant negative effect on purchase intention, supporting hypothesis H4. Value co-creation has a significant positive impact on purchase intention, supporting hypothesis H6. The results of the correlation analysis can provide a preliminary basis for the study hypothesis testing.

4.2. Hypothesis Testing of the Live Characteristics of Fresh Agricultural Products

In this study, the analytical method of structural equation modeling was used [46]. The model-fitting results are shown in

Table 7. Fit Index and Fit Results of the Model.

Fitting Index	X2	df	X2/df	RMSEA	IFI	CFI	TFI
Test Results	556.666	255	2.183	0.062	0.947	0.947	0.937

, which is a good fit and can be used for structural equation modeling.

The structural equation method was applied to correlate the established variables to test the study’s assumptions. In polynomial data analysis, the structural formula is a primary method. It can be used to perform a comprehensive study of the above variables and combine legal factors with measurement models for variables that cannot be directly observed. The metric model refers to the link between latent values and indicators, while in the structural model, it is the interaction between implicit variables. This paper uses these two necessary measures to determine whether the hypothesis is valid.

Figure 2 and

Table 8. Statistical table of path coefficients.

			Path Factor	p
Perceived Risk	<---	Visual	−0.264 ***	***
Perceived Risk	<---	Interactive	−0.226 **	0.001
Perceived Risk	<---	Authentic	−0.189 ***	***
Perceived Risk	<---	Recreational	−0.219 **	0.002
Value Co-creation	<---	Visual	0.202 **	0.004
Value Co-creation	<---	Interactive	0.179 *	0.023
Value Co-creation	<---	Authentic	0.145 **	0.005
Value Co-creation	<---	Recreational	0.162 *	0.033
Consumer Purchase Intention	<---	Visual	0.164 *	0.012
Consumer Purchase Intention	<---	Interactive	0.166 *	0.018
Consumer Purchase Intention	<---	Authentic	0.164 ***	***
Consumer Purchase Intention	<---	Recreational	0.148 *	0.031
Consumer Purchase Intention	<---	Perceived Risk	−0.249 ***	***
Consumer Purchase Intention	<---	Value Co-creation	0.323 ***	***

Note: *, **, *** indicates significance at the 0.05, 0.01, and 0.001 levels, respectively.

show that the live visibility of fresh agricultural products hurts perceived risk. With a significance level of p < 0.001 and a path coefficient of −0.264, hypothesis H1a holds. Live interactivity of fresh agricultural products pulls perceived risk, and with a significance level of p < 0.01 and a path coefficient of −0.226, hypothesis H1b holds. Fresh agricultural products’ live authenticity breaks perceived risk; with a significance level of p < 0.001 and a path coefficient of −0.189, hypothesis H1c holds. Live entertainment of fresh agricultural products hurts perceived risk with a significance level of p < 0.01 and a path coefficient of −0.219, and hypothesis H1d holds. In summary, it is concluded that hypothesis H1: the live characteristics of fresh agricultural products have a significant adverse effect on perceived risk, is maintained.

Live visibility of fresh agricultural products positively impacts value co-creation, with a significance level of p < 0.01 and a path coefficient of 0.202, and hypothesis H2a holds. Live interactivity of fresh agricultural products positively affects value co-creation, with a significance level of p < 0.05 and a path coefficient of 0.179, and hypothesis H2b holds. Live authenticity of fresh agricultural products positively affects value co-creation, with a significance level of p < 0.01 and path coefficient of 0.145, and hypothesis H2c holds. Live entertainment of fresh agricultural products positively affects value co-creation with a significance level of p < 0.05 and a path coefficient of 0.162, and hypothesis H2d holds. In summary, it is concluded that hypothesis H2: the live broadcast characteristics of fresh agricultural produce have a significant positive impact on the value co-creation, is maintained.

Live visibility of fresh agricultural products has a significant positive effect on purchase intention, with a significance level of p < 0.05 and a path coefficient of 0.164, and hypothesis H3a holds. Live interactivity of fresh agricultural products has a significant positive impact on purchase intention, with a significance level of p < 0.05 and a path coefficient of 0.166, and hypothesis H3b holds. Live authenticity of fresh agricultural products has a significant positive effect on purchase intention; with a significance level of p < 0.001 and a path coefficient of 0.164, hypothesis H3c holds. Fresh agricultural products and live entertainment have a significant positive impact on purchase intention, with a significance level of p < 0.05 and a path coefficient 0.148, and hypothesis H3d holds. In summary, hypothesis H3: the live broadcast characteristics of fresh agricultural produce have a significant positive effect on purchase intention, is maintained.

Perceived risk significantly affects purchase intention; with a significance level of p < 0.001 and a path coefficient of −0.249, hypothesis H4 holds. Value co-creation has a significant positive effect on purchase intention with a significance level of p < 0.001 and a path coefficient of 0.323, and hypothesis H6 holds.

4.3. Intermediary Role Relationship Hypothesis Validation

The stepwise regression coefficient analysis modified by Zhonglin Wen and Baojuan Ye (2014) was used to test whether the mediating effect of self-efficacy was valid [47]. The live characteristics of fresh agricultural products are set as X. The assumed perceived risk and value co-creation are designated as M. The purchase intention is set as Y. Three regression equations are established.

$$Y=cX+e_1$$

(1)

$$M=aX+e_2$$

(2)

$$Y=c^{\prime }X+bM+e_3$$

(3)

Step 1: Execute the regression Equation (1) of consumer purchase intention on the live characteristics of fresh agricultural products and test coefficient c. If c is not significant, stop the test and continue to step 2 if it is substantial.

Step 2: Perform the regression Equation (2) of perceived risk and value co-creation on consumer purchase intention and test coefficient a. Stop the test if a is not significant and continue to step 3 if it is substantial.

Step 3: Execute the regression Equation (3) of consumers’ purchase intention on fresh agricultural products live characteristics, perceived risk, and value co-creation, to test coefficient b. If b is not a significant, stop the testl if b is significant, then the tested coefficient, when not necessary, means M plays a fully mediating role. M plays a partially mediating role when powerful and more petite than c. This paper used the Bootstrap method to test the intermediate effect and tested 5000 samples with Bootstrap ML. When 0 is included in 95%, there is no medium effect, and when 0 is not included, the medium impact appears. The final results are shown in Table 8.

As can be seen from the data in

Table 9. Results of intermediate effect test.

Assumptions	Intermediary Path	Indirect Effect Coefficient	95% Confidence Interval
			Upper Limit	Lower Limit
H5a	Visual-Perceived Risk-Consumer Purchase Intention	0.388 ***	0.149	0.438
H5b	Interactive-Perceived Risk-Consumer Purchase Intention	0.349 ***	0.134	0.404
H5c	Authentic-Perceived Risk-Consumer Purchase Intention	0.305 ***	0.108	0.318
H5d	Recreational-Perceived Risk-Consumer Purchase Intention	0.379 ***	0.151	0.432
H7a	Visual-Value Co-creation-Consumer Purchase Intention	0.450 ***	0.108	0.353
H7b	Interactive-Value Co-creation-Consumer Purchase Intention	0.460 ***	0.106	0.368
H7c	Authentic-Value Co-creation-Consumer Purchase Intention	0.348 ***	0.075	0.245
H7d	Recreational-Value Co-creation-Consumer Purchase Intention	0.440 ***	0.101	0.343

Note: *** indicates significance at the 0.001 level.

, none of the eight hypotheses tested contained 0 in the 95% confidence interval. This indicates a mediating effect of perceived risk and value co-creation between fresh agricultural produce characteristics and purchase intention.

5. Discussion

Through the above structural equation model analysis and stepwise regression coefficient analysis, it is shown that the visibility, interactivity, authenticity, and entertainment of fresh agricultural products live broadcast all have a significant positive impact on consumers’ purchase intention and perceived risk. Value co-creation plays a mediating role between the characteristics of fresh agricultural products and purchase intention.

Hypothesis H1, that the live characteristics of fresh agricultural products have a significant adverse effect on perceived risk, is maintained. Among them, visibility and authenticity are significant at the 0.001 level. In traditional fresh agricultural produce e-commerce, due to the bias of “seller show” and “buyer show,” the risk of goods increased. The introduction and application of the anchor and direct observation of the product can significantly reduce consumers’ perception of risk. The professionalism of the anchor is the basis for consumers to form trust in the face of numerous redundant information and accurate answers to various questions about the purchase and transportation process. Hosts need to be clear about the different details in this series, from the generation of fresh agricultural produce to the final product to the consumer, and their own choice of high-quality goods, which will increase consumer trust and thus reduce consumer perceived risk.

Hypothesis H2, that the live broadcast characteristics of fresh agricultural products have a significant positive effect on value co-creation, is maintained. The magnitude of the influence of the four fresh agricultural produce live features on value co-creation is: live visibility in fresh agricultural products > live entertainment in fresh agricultural products > live interactivity in fresh agricultural products > live authenticity in fresh agricultural products. The distinction in significance levels shows that visibility and authenticity have the most significant impact on value co-creation. Consumers receive much visual information in the live broadcast. The live format to bring consumers the real sense will make it easier for consumers and businesses to form unified values, increase consumer feedback on suggestions with companies, and form a reasonable exchange of information. Timely interaction enables the anchor to instantly grasp the needs of consumers and adjust the program’s content promptly, enhancing consumer engagement. During the live broadcast of fresh agricultural products, the focus should be on the live questions from consumers to ensure that consumers’ questions are answered during the broadcast. Please pay attention to whether the fresh goods you see on the live broadcast are the same as those you buy to ensure their authenticity. It should focus on optimizing the scene and user interface of the e-commerce platform to enhance the comfort of the live visibility of the fresh agricultural produce of the product. The anchor should focus on the entertainment of the live broadcast so that the consumers can have fun with the purchase. Hosts can take a more personal approach to users and answer questions from customers promptly in messages to enhance user interactivity.

Hypothesis H3, that the live characteristics of fresh agricultural products have a significant and positive effect on consumers’ purchase intention, is maintained. Authenticity is the most critical factor in forming consumers’ purchase intentions. This result is inconsistent with Zhang Baosheng’s (2021) research on the characteristics of live broadcasts on the Internet about consumers’ purchase intentions, where scholars researched unified live e-commerce. Entertainment was the main factor influencing consumers’ purchase intentions [48]. Consumers will choose to buy products for electrical appliances, makeup, and other e-commerce live broadcasts because of the atmosphere of the live broadcast room and the entertaining nature of the anchor. In fresh agricultural produce live broadcast, consumers choose to buy products based on the authenticity of the products in the live broadcast. This shows differences between fresh agricultural produce live broadcast and live broadcast of other categories, and there is a need to study them separately.

Hypothesis H4, that perceived risk has a significant negative effect on purchase intention, and hypothesis H6, that value co-creation has a significant positive impact on purchase intention, are both maintained. When Yuan Liang et al. (2017) studied uncategorized e-commerce, they concluded an adverse effect of perceived risk on purchase intention [26]. From the data in this paper, it is seen that the conclusion that perceived risk has a significant negative impact on purchase intention still holds for live fresh agricultural produce. The significant positive effect of value co-creation on purchase intention indicates that value co-creation is essential in increasing consumers’ purchase intention.

Hypothesis H5 is that perceived risk mediates between fresh agricultural product live broadcast characteristics and purchase intention. Hypothesis H7 is that value co-creation mediates both fresh agricultural product live broadcast characteristics and purchase intent. Perceived risk and value co-creation play a mediating role between the live parts of fresh agricultural products and purchase intentions. The findings illustrate that when consumers judge whether to purchase fresh agricultural products in the fresh agricultural live broadcast, consumer preferences and choices change due to changes in perceived risk and value co-creation. This paper argues that the current research focuses on avoiding perceived threats to consumers in fresh agricultural live. Still, this paper argues that preventing perceived risk is as important as improving value co-creation.

6. Conclusions

Based on the results of this paper, the following conclusions can be drawn. First, the live feature of fresh agricultural produce will have a significant positive effect on consumers’ willingness to buy, and the authenticity of fresh agricultural produce is an aspect that consumers attach more importance to. Second, the live characteristics of fresh agricultural products significantly affect perceived risk and value co-creation, with authenticity having the most significant impact on perceived risk and visibility having the most excellent effect on value co-creation. Third, perceived risk and value co-creation play a mediating role between the live characteristics of fresh agricultural products and purchase intentions.

Combining the characteristics of live broadcast fresh agricultural products, the following suggestions are made for live broadcast merchants: Improve the income and competitive advantage of fresh agricultural produce live broadcast merchants to promote the sustainable and healthy development of fresh agricultural produce live broadcast. In more detail:

(1) Enrich anchor expertise to enable consumers to reduce perceived risk. (2) Choose more agricultural products production areas or breeding bases, etc., which promote the authenticity of the live background place. This will bring consumers a deeper understanding of the actual situation of the origin and show them more real planting or breeding scenes, which will help enhance consumers’ perception of the authenticity of fresh agricultural products broadcast. (3) Focus on the needs of consumers and highlight their concerns. (4) In contrast to traditional value co-creation behavior, consumer participation in value co-creation implies recognition of the company’s values. In an online environment, consumer engagement will increase, but not all consumers will agree with a company’s values. In this context, live fresh food e-commerce must take the initiative to put forward its value proposition to gain more consumer recognition.

The shortcomings of this paper are mainly reflected in the following two points. First, in this paper, in exploring the influence of consumers’ purchase intention of live fresh agricultural products, the perceived risks are not classified in detail due to space limitation, leading to perceived risks not being more refined. Second, a large sample was not collected due to workload constraints, and the study results may change if the sample size is expanded.