Examining the influence of technology-enhanced information on purchase of green food in the context of live streaming commerce

ABSTRACT

With the growing concerns about ecological pressure and food safety, researchers and marketers have devoted to finding effective ways to bridge the gap between consumers’ high interest and weak purchasing behaviour towards green food. The unique information advantages of live streaming commerce help consumers improve their cognition of green food’s benefits, evoke their concerns about health and environmental issues, and thus form positive purchasing behaviour. This study develops a conceptual framework by incorporating health consciousness and environmental awareness as mediating variables to estimate how live streaming information features impact consumers’ purchasing intention and behaviour towards green food. Considering the important influences of streamer expertise and consumer attitude on purchasing behaviour, the two variables are employed in our model as moderating variables. Structural equation modelling is used to analyse data collected from 948 Chinese live streaming shoppers. The results reveal that information richness, authenticity, synchronicity, and sociability significantly enhance consumers’ health consciousness and environmental awareness, and the favourable purchasing intention as well as behaviour is the result of combined impacts of health consciousness and environmental awareness. Both streamer expertise and consumer attitude play important roles in promoting consumers who intend to purchase green food to become actual buyers. The integrated model expands the empirical research of green food consumption from the perspectives of specific marketing patterns, seller and consumer characteristics, and provides a broader insight for researchers, sellers, and policy-makers to understand green food consumption and supply in emerging economies.

KEYWORD:

• Information feature
• health consciousness
• environmental awareness
• purchasing behaviour
• green food

1. Introduction

Consumers’ dietary choices have been much concerned both in developed and developing countries (Eyinade et al., 2021). With the rapid economic development and urbanization of China, on the one hand, the magnitude and preferences of Chinese food consumption have changed greatly; on the other hand, serious environmental pollution has raised people’s concern about food safety (Liu et al., 2022; Tong et al., 2020). It is reported that food sales revenue exceeded 20 billion yuan, and the net profit reached 1.21 billion yuan in 2022 (China Food Limited, 2023). Although the Chinese green farming practices have kept an increasing trend over the past decade (Chu, 2018; Tong et al., 2020), the per capita consumption of green food has remained at a relatively lower level. Specialists and researchers indicate that green food industry has great development potentials, green food consumption becomes prevalent and affordable with the application of new technologies (Asian et al., 2019). In the context of the global COVID-19 pandemic, the quality and safety of food are widely concerned. Consumers believe that green food can improve immunity, prevent various chronic diseases, and thus bolster health (Grinberga-Zalite et al., 2021; Parashar et al., 2023). The increasing food safety concern and health consciousness drive people to pay more attention to the healthier and safer food, such as green food, while relatively few people actually put it into practice (Qi & Ploeger, 2021; M. Yang et al., 2023). The gap between consumers’ high attention and weak purchasing behaviour might be caused by uncertainty about product quality (Loebnitz & Grunert, 2018). The market information imperfection and asymmetry are essential reasons for this phenomenon. Beyond that, relatively less advertising leads to a lower per capita consumption level (Chandrashekhar, 2014).

The rapid development of E-commerce promotes the online sale and shopping of a variety of products. The number of online food shoppers maintains an upward trend (The 51st statistical report on China Internet Network Information Center, 2022), and numerous green food sellers develop and expand the online business to seize new market opportunities. Although E-commerce facilitates consumer accessibility of information and products, it is difficult for consumers to distinguish or identify green food products based on their own experience. Improving consumers’ cognition of green food’s benefits and building trust are crucial prerequisites for agricultural companies and sellers to survive in the online fierce environment (Chen et al., 2018; Lindström et al., 2022; Loebnitz & Aschemann-Witzel, 2016). However, an enormous amount of information, exaggerated literal statements, and processed pictures or videos increase the difficulty of consumer decision-making. Indistinguishable information and virtual trading space raise shopping risks and impede consumers’ enthusiasm for buying green food online. The problem of information asymmetry still exists in the online business (Mavlanova et al., 2016), which means that traditional online marketing patterns may not be suitable for green food products (Neumayr & Moosauer, 2021).

Recently, a new commerce pattern, live streaming commerce, emerges and quickly attracts wide attention due to its prominent effect on promoting product sales (Rungruangjit, 2022). During 2022 “Singles Day” shopping festival (October 31st 20:00-November 11th 23:59), the sales of live streaming commerce reached 181.4 billion yuan. Most E-commerce and social media platforms in China provide live streaming services to online users and take them as important measures to raise revenue. The users of live streaming commerce exceeded 515 million in China by December 2022 (The 51st statistical report on China’s Internet development, 2022). It becomes an increasingly prevalent shopping medium, because it is not only a novel online shopping way, but also a better way to acquire abundant and effective information and knowledge with a lower cost, and help people kill boring time (Gao et al., 2021; M. Zhang et al., 2022).

The previous study presents that live streaming commerce greatly promotes the sales of agricultural products during novel coronavirus outbreaks (Peng et al., 2021). In September 2022, “Taobao” live streaming platform conducted 200,000 village live streaming activities, attracting more than 700 million viewers and driving 4 million orders (The 51st statistical report on China’s Internet development, 2022). Compared with conventional food products, consumers are more cautious about buying green food due to the higher price premium and quality uncertainty. Helping consumers to be more informed and aware of green food traits will contribute to reducing quality uncertainty and increasing their purchasing intention and behaviour.

The signalling theory has been widely applied in economics literature to estimate how the different types of signals that sellers provide affect buyers’ evaluation of product quality (Kirmani & Rao, 2000; Lu & Chen, 2021; Mavlanova et al., 2016). The existing research with respect to product quality uncertainty uses the signalling theory to explain the impacts of information signals on consumers’ perceptions and attitudes towards products and sellers (Lu & Chen, 2021). The noticeable and prominent advantages of live streaming commerce, such as the abundant, pertinent, and reliable information, vivid and actual display, and higher consumer involvement, enable it to convey more effective signals to consumers, which will greatly improve consumers’ cognition of green food’s benefits, reduce uncertainty, and evoke positive psychological responses. Green food is considered to be more nutritious, healthier, and environmentally friendly (Parashar et al., 2020). Thus, a better understanding of green food’s benefits to health and the environment will develop and enhance consumers’ health consciousness and environmental awareness, which are proven to be essential drivers of green food consumption (Eyinade et al., 2021; Parashar et al., 2023). Live streaming marketing is considered to be an appropriate pattern for expanding the online business of green food, while to the best of our knowledge, there is a lack of literature exploring green food purchasing behaviour as well as the determinants in the context of live streaming commerce.

As streamers become the “opinion leaders” of live streaming channels, their personal characteristics play an important role in guiding consumers’ demand and stimulating shopping. The sales-promoting influence of celebrity streamers is especially evident due to their huge number of followers (Ladhari et al., 2020; Ma, 2021b; Rungruangjit, 2022). Unlike apparel and cosmetic products, consumers’ online purchase of green food tends to be more rational and prudent. Therefore, streamer expertise might become one important influencing factor with respect to green food sales. As studies posit that consumer attitude is an important predictor of purchasing intention and behaviour towards environmentally friendly food (Lee & Yun, 2015; Parashar et al., 2023; Qi & Ploeger, 2021), thus this study also explores the influences of attitude and streamer expertise on consumers’ purchase of green food.

Understanding and estimating the features of live streaming commerce and their potential impacts on consumers’ psychological responses as well as purchasing behaviour, will broaden the theoretical research of the green food consumption behaviour and help agricultural enterprises and sellers better understand consumer demand and key factors influencing their online purchasing behaviour, thus formulating appropriate marketing strategies. The main contributions of the study are as follows: (1) we shed light on five prominent information features of live streaming commerce associated with green food marketing; (2) propose a conceptual model to explore the impacts of information features on purchasing intention and behaviour towards green food by incorporating consumers’ health consciousness and environment awareness; (3) estimate the mediating effects of health consciousness and environment awareness to verify the theoretical mechanism as well as the causal relationships among the constructs; (4) further explore the relationship between purchasing intention and purchasing behaviour by employing consumer attitude and streamer expertise as moderating variables; (5) finally provide the theoretical implications and practical contributions based on research results.

2. Literature review

2.1. Green food purchasing behavior

In a great measure, the reason why consumers choose to buy green food is that it is much healthier than conventional food (Dangelico et al., 2021; Loebnitz & Aschemann-Witzel, 2016; Ran et al., 2022; Wang et al., 2019). Green food choices reflect a health-oriented lifestyle (Lazzarini et al., 2018), and consumers’ health consciousness drives them to purchase natural and nutritious food (H. V. Nguyen et al., 2021; Parashar et al., 2023; Qi & Ploeger, 2021). Beyond that, the non-functional attributes of green food, such as the aroma, flavour, appearance as well as environmentally friendly traits, also are regarded as important motives for purchasing green food (Das et al., 2020; Parashar et al., 2020; Sadiq et al., 2021). Consumers begin to realize the benefits of green food to the environment in the context of increasing ecological pressure (D. T. Nguyen & Truong Dt, 2021; Parashar et al., 2020). With the increase in health consciousness and environmental awareness, consumers prefer to choose nutritious and sustainable food products (S. Li & Jaharuddin, 2021; Parashar et al., 2023).

Consumers’ cognition and trust in nutritious and environmentally friendly attributes of green food help to evoke positive psychological evaluations and promote the formation of purchasing intention and behaviour (Qi & Ploeger, 2021; Tong et al., 2020). However, the favourable cognition and trust are difficult to establish because the nutritious and natural contents as well as environmentally friendly production process can not be detected by consumers. Theoretically, one method to strengthen consumers’ cognition and trust is to use Eco-labels approved by third parties like certificate authorities (Liu et al., 2022; Neumayr & Moosauer, 2021; Rihn et al., 2019). For example, the Chinese government labels safe and harmless food products officially as “Organic food, Green food or Pollution-free food”. These Eco-labels are regarded as the signals of attribute merits and usually use logos to quickly attract consumers’ visual attention (Rihn et al., 2019). Hence, Eco-labels are expected to generate positive psychological cues for consumers. In fact, these labels do not exert much impact on consumers’ food choices in practice. This is probably due to the complexity and diversity of Eco-labels, resulting in consumers’ confusion, distrust, and dissatisfaction. The lower public cognition of Eco-labels largely discourages consumers from purchasing the certified food (Loebnitz & Aschemann-Witzel, 2016). Therefore, the government’s efforts for gaining consumer trust and improving cognition of environmentally friendly food by using Eco-labels tend to be lower efficient.

Practitioners and researchers study alternative schemes of Eco-labels with respect to strengthening cognition and trust. Community supported agriculture (CSA) is known as an effective pattern to boost consumer trust and cognition of green food by attracting people to visit rural farms and engage in agricultural practices (Medici et al., 2021). Consumers’ cognition of healthy and environmental benefits of green food will be improved in the participating process, thus leading to a positive attitude towards purchasing green food. However, the CSA pattern has several prerequisites: people should have positive engagement intention and sufficient time and energy to get involved in agricultural practices. Nowadays, the changes in the socioeconomic environment aggravate people’s work pressure, and their spare time is more fragmented, especially for young people. Moreover, the CSA farms are usually far away from the cities, and the sanitary and safety conditions of some CSA farms tend to be not satisfactory either. Consumer involvement in agricultural practices is facing various challenges, and the positive influences of CSA pattern are not as good as expected in China. Therefore, it is urgent to innovate the pattern of consumer involvement and make it more compatible with people’s daily life to enhance consumers’ cognition and trust in green food.

2.2. Live streaming

The existing literature explores the antecedents of live streaming users’ watching intention (Chen & Lin, 2018; Hu et al., 2017), users engagement (Hilvert-Bruce et al., 2018; Wongkitrungrueng & Assarut, 2020), usages and continuous use intention (Y. Li et al., 2021; Sjöblom & Hamari, 2017; M. Zhang et al., 2022; Singh et al., 2021), users’ satisfaction with live streaming shopping (Ma, 2021a), and some specific behaviours, such as gift-giving behaviour (Yu et al., 2018). Some studies estimate the predictors of users’ purchasing intention in the context of live streaming commerce. Sun et al. (2019) estimate the impacts of visibility, meta voicing, and guidance shopping on customer purchase intention from the perspective of IT affordance. Ma (2021b) states that an individual’s shopping intention is motivated by gratification acquired from perceived enjoyment, social interaction, social presence, and utility. Gao et al. (2021) apply the Elaboration Likelihood Model to examine the effects of the central route factors (information completeness, accuracy, and currency) and peripheral route factors (streamer trustworthiness, attractiveness, bullet-screen consistency, and co-viewer involvement) on viewers’ perceived persuasiveness, and illustrate that both route factors significantly influence viewers’ perceived persuasiveness as well as purchasing intention. Lu and Chen (2021) focus on clothes and cosmetics live streaming and estimate the influences of streamers’ characteristics and values similarity on reducing product uncertainty and establishing trust as well as purchasing intention. Park and Lin (2020) discuss the influences of various matches with internet celebrities on consumer attitude as well as purchasing intention. Huang and Suo (2021) illustrate four factors that might drive the impulse buying decision of live streaming viewers and estimate their casual relationships by using perceived risk as a mediating variable. Table 1 shows the details of the relevant studies.

Table 1. Literature review.

Authors and year	Independent variables	Intermediate variables	Outcome Variables	Mediators/moderators	Methods	Samples	Country
Hilvert-Bruce et al. (2018)	entertainment, information seeking, meeting new people, social interactions, social support, sense of community, social anxiety, external support		Live streaming engagement (emotional connectedness, time spent, time subscribed, donations)		Multiple and ordinal linear regression	Twitch users (2227)	Australia
Wongkitrungrueng and Assarut (2020)	Utilitarian value, hedonic value, symbolic value	Trust in products, trust in sellers	Customer engagement	Mediators: trust in products, trust in sellers	Partial least squares structural equation modeling (PLS-SEM)	Viewers of Facebook (261)	Bangkok, Thailand
Sjöblom and Hamari (2017)	Motivators (affective, cognitive, personal integrative, social integrative, tension release)		Usages (hours watched, streamers watched, streamers followed, subscription)		Component-based PLS-SEM	Twitch users (1097)	Australia
Y. Li et al. (2021)	Interaction, identification, synchronicity, Vicarious expression	Emotion attachment to streamers, platform attachment	Visit duration, user retention		Structural equation modeling
Singh et al., 2021	Effort expectancy, performance expectancy, convenience value, Monetary value, emotional value, social value, perceived entertainment, perceived risk, addition	Personal innovativeness, perceived value	Continuous use intention		Partial least squares +artificial neural network (PLS+ANN)	Streaming services user (869)	India
M. Zhang et al. (2022)	Active control, synchronicity, two-way communication, personalization, visibility	Trust in products; Trust in streamers	Continuous use intention	Mediators: trust in products, trust in streamers Moderators: live streaming genre	Structural equation modeling	“Taobao” Live Streaming users (446)	China
Chen and Lin (2018)	Flow, entertainment, social interaction, endorsement	Attitude, perceived value,	Watching intention	Grouping variables: gender, age	Structural equation modeling	Live streaming viewers（313）	China
Hu et al. (2017)	Experience of parasocial interaction, Actual self-congruity, Ideal self-congruity, participation, Cognitive communion, Resonant contagion	Broadcaster identification, group identification,	Continuous watching intention	Moderators: live streaming genres	Partial least squares (PLS)	Live streaming audiences（412）	China
Yu et al. (2018)	Viewer engagement (chat and recommendations)		Gift-giving behaviors		Binary logistic regression and negative binomial regression	Viewers of “Afreeca TV” (2294837)	South Korea
Ma (2021a)	Source expertise, media richness	Information quality, service quality, argument quality,	satisfaction	Mediators: social presence, trustworthiness	Structural equation modeling	Live-stream shoppers (237)	China
Sun et al. (2019)	Visibility, meta voicing, guidance shopping	Immersion, presence	Purchase intention	Mediators: immersion, presence	Partial least squares (PLS)	Live streaming shopping users (510)	China
Ma (2021b)	Perceived network size	Perceived enjoyment, perceived social interaction, social presence, perceived utility, self-presentation	Live shopping intention	Mediators: Perceptions of digital celebrities, experience-focused shopping	Structural equation modeling	Consumers of live streaming shopping (339)	China
Gao et al. (2021)	information completeness, accuracy, and currency, streamer trustworthiness, streamer attractiveness, bullet-screen consistency, co-viewer involvement	Perceived persuasiveness	Purchasing intention, Response intentions	Moderators: Mindfulness	Structural equation modeling	Live streaming shopping users（395）	China
Lu and Chen (2021)	Physical characteristic similarity; value similarity	Product fit uncertainty, product quality uncertainty, trust,	Purchase intention		Structural equation modeling, multi-group analysis, and qualitative case analysis	Live streaming shopping users（535）	China
Park and Lin (2020)	Wanghong-product fit, live content-product fit, Self-product fit	Wanghong trustworthiness, wanghong attractiveness, Utilitarian attitude, hedonic attitude	Intention to buy		Structural equation modeling	Live streaming viewers (252)	China
Huang and Suo (2021)	Price promotion, promotion time limit, perceived opportunity cost, streamer-consumer interaction, consumer-consumer interaction, visual appeal	Perceived risk	Impulse buying decisions	Mediators: perceived risk	Structural equation modeling	Live streaming shopping users（535）	China

Live streaming commerce as a new and unique business model quickly attracts online shoppers’ attention due to its real-time interactivity, involvement, entertainment, and novelty (Hu et al., 2017; Yu et al., 2018; F. Zhou et al., 2019). Different from static pictures and text-based product descriptions on E-commerce websites, the abundant audio-visual information provided by live streaming services is more likely to arouse consumers’ interests and increase stickiness (F. Zhou et al., 2019). It can provide consumers with more complete and detailed product information from multiple dimensions to eliminate their shopping risks and help them make purchasing decisions (Wongkitrungrueng & Assarut, 2020). The real-time interaction environment makes it easier for consumers to generate a strong sense of involvement (Hilvert-Bruce et al., 2018). In this context, abundant information, high consumer involvement, and intensive interaction are conducive to better understanding the benefits of green food, thus driving viewers’ positive psychological responses as well as purchasing behaviour.

As live streaming commerce is regarded as one suitable medium for green food trades, the mechanism and routes that the unique features of live streaming commerce impact consumers’ purchasing intention and behaviour towards green food are worth studying. However, limited research refers to this topic. Because the commerce patterns, consumers’ lifestyles, people’s health consciousness, and environmental awareness might be different between developing countries and developed countries, this study would help to understand the determinants of green food purchasing behaviour in emerging economies.

3. Conceptual model

3.1. Model framework

Signals are usually regarded as the features of an object (Kirmani & Rao, 2000). When consumers use live streaming shopping, the unique information features of live streaming commerce contribute to conveying more signals to viewers, thus helping viewers form positive cognitive evaluations of product quality (Sun et al., 2019). The key parameters that this study focuses on are the information features that consumers can not experience from traditional online shopping websites or the more prominent features. In order to gain a better understanding of the information features of live streaming commerce, we conducted a literature review to identify the potential features. Compared with traditional online shopping websites, live streaming commerce is able to convey abundant, detailed, dynamic, and visual-audio information (Lu & Chen, 2021; Ma, 2021b; X. Zhang et al., 2019), authentic, correct, and reliable information (Sun et al., 2019; Yu et al., 2018), synchronous, immediate, and instantaneous information (Chen & Lin, 2018; Hu et al., 2017; Wongkitrungrueng & Assarut, 2020; F. Zhou et al., 2019), and personalized information (Sun et al., 2019; H. Yang & Lee, 2018; M. Zhang et al., 2022). The information will greatly reduce consumers’ uncertainty about green food benefits and facilitate their purchasing decisions. In addition, live streaming commerce enables viewers to interact with streamers as well as other viewers (Sun et al., 2019; Zhao et al., 2018). The social attribute of live streaming commerce is considered to be a significant determinant influencing viewers’ loyalty (Hsu et al., 2020). It helps to build friendships between viewers and streamers (Y. Li et al., 2021). The previous research indicates that social media activities play an important role in forming purchasing decisions of today’s generation (Agrawal et al., 2023). Lu and Chen (2021) illustrate that live streaming commerce increases viewers’ purchasing intention through product and social routes. The product route mainly conveys utilitarian information to reduce product fit uncertainty. The social route mainly conveys emotional information to build intimate relationships between streamers and viewers, which lead to a high level of trust (M. Zhang et al., 2022). In the context of Chinese culture, face consciousness and group conformity are considered as important factors influencing consumers’ purchasing intention towards green food (Qi & Ploeger, 2021), so the impact of socialization information on consumers’ psychological responses as well as purchasing behaviour is discussed in this study. Based on the above analysis, we finally identify five information features of live streaming commerce, including information richness, authenticity, synchronicity, personalization, and sociability from utilitarian and affective standpoints.

The external environmental cues affect the formation of consumers’ environmental awareness, so educating consumers on green products will evoke positive environmental attitudes, increase environmental awareness, and then generate desirable behavioural intentions (Agrawal et al., 2023; Majerova et al., 2020; Taufique et al., 2016). Researchers illustrate that consumers’ environmental awareness plays a vital role in green consumption decisions (Chu, 2018; Xu et al., 2020; M. Yang et al., 2023). Further, the outbreaks of COVID-19 arise people’s concern of health issues. The information and knowledge with respect to green food’s nutritious, natural, and safe contents will drive consumers’ concern for health state, enhance health consciousness, and then lead to the favourable purchasing intention and behaviour towards green food. Consumers’ health consciousness is identified as a significant determinant of organic and green food consumption (Chu, 2018; Nagaraj, 2021; Parashar et al., 2023; Wang et al., 2019). Making viewers aware of health and environmental issues through rich, authentic, synchronic, personalized, and social information will drive favourable purchasing intention and behaviour towards green food.

The stimulus-organism-response framework (SOR) (Mehrabian & Russell, 1974) enables researchers to examine the impacts of external information signals on consumers’ psychological responses as well as behavioural intention (Agrawal et al., 2023; Lee & Yun, 2015), so it is employed as a theoretical foothold. In this study, information features are defined as the external stimuli that affect consumers’ internal states, and consumers’ health consciousness and environmental awareness are employed as the proxies of consumers’ internal states to predict purchasing intention and behaviour. Finally, we develop a conceptual framework by incorporating health consciousness and environmental awareness to estimate how information features of live streaming commerce impact consumers’ purchasing intention and behaviour towards green food. The model framework is described in Figure 1. In addition, streamer expertise and consumer attitude are included in this model as moderating variables.

Figure 1. Conceptual framework of the theoretical model.

3.2. Hypothesis development

3.2.1. Information richness

Information richness refers to the degree of providing product information from multiple dimensions in many ways (M. Zhang et al., 2022). The visual-audio features of live streaming commerce enable viewers to obtain plenty of information to facilitate their purchasing decision-making (Wongkitrungrueng & Assarut, 2020). Previous literature illustrates that the product information acquired from live streaming services is richer and more comprehensive (Gao et al., 2021; Neumayr & Moosauer, 2021; Sun et al., 2019). Live streaming services help viewers get a piece of much greater overall information about green food products. On the one hand, streamers elaborate on the health-related benefits of green food, such as the natural, nutritious, safe traits, etc. On the other hand, streamers display green food products from various angels, and cook and savour the food products in live streaming channels. Streamers’ descriptions of the appearance, colour, smell, and tasty release good product quality signals, leave vivid product images to viewers (Gao et al., 2021; Sun et al., 2019), and reduce viewers’ doubts about product quality (M. Zhang et al., 2022; L. Zhou et al., 2018). Streamers also share consumption experiences and tricks, novel recipes, and scientific cooking methods to arouse viewers’ curiosity and interests. The rich, comprehensive, and dynamic information is more persuasive than static text and picture information (Ma, 2021a). It enables viewers to comprehensively understand the product characteristics, leading to positive cognitive attitudes and behavioural intentions (J. W. Kang & Namkung, 2019; M. Zhang et al., 2022). The detailed and rich information will help viewers improve the cognition of nutritious and safe traits of green food and acquire more health-related knowledge, which will drive viewers to reflect on their health status, and thus develop and enhance their health consciousness. Consumers’ health consciousness is one of the important motivating factors for green food consumption (Lee & Yun, 2015; Qi & Ploeger, 2021).

Many countries are taking initiatives to encourage sustainable agricultural practices (Agrawal et al., 2023; Chu, 2018). Numerous agricultural corporations develop and apply environmentally friendly production technologies. Meanwhile, marketers are also taking initiatives to increase consumers’ cognition of green practices (Liu et al., 2022; Uzun & Keles, 2012) and attempting to build a stronger market share. Advertising green food through the live streaming medium is an effective way to strengthen this cognition. Streamers highlight the policy orientation and producers’ green practices, which will increase viewers’ cognition of the altruistic attributes of green food. The environment-related knowledge evokes people’s concern about environmental issues, increases their environmental awareness, and promotes green behaviours (Wang et al., 2019; R. Yang et al., 2020). The rich information about green food that viewers receive from live streaming services is conducive to increasing their health and environment-related knowledge, and developing and enhancing their health consciousness and environmental awareness, which finally affect the choices of green food. Based on the above analysis, the following hypothesis is proposed.

H1.

Information richness will positively impact viewers’ health consciousness (H1a) and environmental awareness (H1b).

3.2.2. Information authenticity

Information authenticity is defined as the degree of providing authentic, accurate, and credible information (Gao et al., 2021). It is considered as a crucial factor influencing viewers’ acceptance of the products (J. W. Kang & Namkung, 2019). The information provided by videos is more authentic than text (Sundar & Limperos, 2013). Xu et al. (2021) present that the images of live streaming products are affected by streamers’ characteristics. Streamers, as the main information sources, play a significant role in the viewers’ decision-making process (Park & Lin, 2020). Some research indicates that streamer expertise is not considered to be associated with information quality (Gao et al., 2021; Park and Lin, 2020). Viewers think that the information is authentic and credible, which are probably because viewers are attracted by streamers’ special talents or personal charisma (Hu et al., 2017), or even the professionalism in other domains. This is the case that agricultural corporations prefer to cooperate with celebrity streamers even if they are just professional in the domains of cosmetics or apparel. Viewers are more likely to be loyal fans of attractive streamers. They tend to believe that the information provided by these streamers is authentic, and the recommended products are suitable for them, thus generating positive purchasing intention and behaviour (Lu & Chen, 2021; Qi & Ploeger, 2021; Tong et al., 2020).

Consumers always seek authentic rather than artificial shopping experiences (Ma, 2021a). Compared with traditional E-commerce websites, consumers can observe the actual green food products rather than the over-beautification pictures or videos (M. Zhang et al., 2022), and viewers can observe the true reactions of streamers when they cook and savour food products. Streamers are considered as reliable information sources because they are less influenced by corporate interests (Ma, 2021a). The authentic and reliable information helps streamers develop positive word of mouth and increase viewers’ loyalty. Therefore, streamers are less likely to convey false and ambiguous information to viewers. Meanwhile, streamers also use objective materials, such as videos of the production process, authoritative reports and labels, etc., to convey reliable health and environment-related information to viewers. Streamers invite agricultural experts, local officers, or celebrities to come to live streaming channels to introduce products, these activities release a signal of acceptance by others or third parties, and thus further strengthen the information authenticity. The accurate and unambiguous information can reduce the potential risks caused by misunderstanding (Gao et al., 2021), improve viewers’ cognition of green food’s benefits to health and the environment, and make viewers believe that consuming green food is appropriate for improving physical health and reducing environmental pressure. Based on the above analysis, the following hypothesis is proposed.

H2.

Information authenticity will positively impact viewers’ health consciousness (H2a) and environmental awareness (H2b).

3.2.3. Information synchronicity

Information synchronicity refers to the degree of providing instantaneous information in a real-time interaction manner (Hou et al., 2019; Y. Li et al., 2021). Real-time and highly intensive interaction is considered as a salient difference between live streaming shopping and traditional websites shopping (M. Zhang et al., 2022). Live streaming services allow viewers to directly interact with streamers or other viewers in a real-time manner (Holt et al., 2016; Zhao et al., 2018). Viewers are able to directly ask questions and post comments with respect to products, services, or some other topics unrelated to the products without converting the page (Sun et al., 2019), which fulfil consumers’ certain psychological needs, such as seeking information and self-expression (Chen & Lin, 2018; Yu et al., 2018). Streamers respond viewers’ comments and questions quickly and provide help and suggestions to viewers (Hu et al., 2017), giving a signal of sincerity to viewers. The real-time interaction increases consumer satisfaction (Ma, 2021a; Rhee & Choi, 2020) and consumer stickiness (X. Zhang et al., 2019), leading to positive cognitive evaluations. The synchronous interaction enables viewers to experience efficient communication (M. Zhang et al., 2022), which helps to increase viewers’ knowledge of green food’s benefits to health and the environment, making them more conscious of health and environmental issues.

Synchronous interaction expands the width and depth of information transmission. Viewers can not only obtain immediate green food information from streamers, but also obtain a large of synchronous information through real-time bullet screens interaction (Hou et al., 2019; F. Zhou et al., 2019). The co-viewers’ comments, opinions, and questions might broaden viewers’ insights, and enable viewers to get more comprehensive health and environment-related information that they might not think of (Sun et al., 2019). The information expressed by many others is considered to be more reliable (Gao et al., 2021), and viewers are more likely to accept and adopt it. Moreover, viewers can obtain shopping and using experiences from co-viewers. Co-viewers’ purchasing experiences and referrals are considered as strong signals about product quality (Xu et al., 2017), which contribute to strengthening viewers’ cognitive evaluations of green food’s benefits. The more synchronous information viewers obtain, the more likely they are to raise concerns about health and environmental issues, and evoke health consciousness and environmental awareness. Based on the above analysis, the following hypothesis is proposed.

H3.

Information synchronicity will positively impact viewers’ health consciousness (H3a) and environmental awareness (H3b).

3.2.4. Information personalization

Information personalization refers to the degree of providing personalized and customized information (Sun et al., 2019; Wongkitrungrueng and Assarut, 2018). The personalization marketing strategy plays an important role in the success of E-commerce (Kang et al., 2020; Ma, 2021b; Rhee & Choi, 2020). Information can change consumers’ behaviour effectively when it is tailored to consumers’ full shopping journey (Ran et al., 2022). Live streaming services enable viewers to obtain the information they want in many ways. When consumers use live streaming shopping, they can ask streamers to display the products they are interested in and autonomously decide how to interact with them (M. Zhang et al., 2022). Viewers also can bring forward the desired products that are not linked in live streaming channels, or ask questions, seek purchasing and using suggestions based on their personal needs (Hu et al., 2017; Yu et al., 2018; F. Zhou et al., 2019). Streamers encourage viewers to express their interests, concerns, and requirements through bullet screens or product comment areas and meantime provide feedback and guidance, which help consumers solve individual problems, leading to positive cognitive evaluations as well as purchasing intention and behaviour (Huang & Suo, 2021; Sun et al., 2019; Yu et al., 2018).

Personalized services meet consumers’ specific needs and facilitate them to find valuable information about their desirable products (Ma, 2021b). It greatly improves viewers’ perception of information usefulness and promotes the formation of positive psychological responses, such as satisfaction (Ma, 2021a). Highly personalized communication gives viewers a signal of being taken seriously and makes them realize that streamers are in their best interests (M. Zhang et al., 2022). Consumers are more likely to be persuaded if the information is provided on the basis of their personal interests and concerns (Ma, 2021a). Such personalized information is more pertinent (Gao et al., 2021) and greatly reduces the product fit uncertainty. It contributes to convincing viewers that consuming green food is appropriate in terms of improving physical health and alleviating ecological pressure. The perceived personal relevance of health-related information drives consumers to purchase green food products (Loebnitz & Grunert, 2018). Personalization information is conducive to stimulating viewers’ concerns and potential needs by increasing their cognition of green food’s benefits to health and the environment. Based on the above analysis, the following hypothesis is proposed.

H4.

Information personalization will positively impact viewers’ health consciousness (H4a) and environmental awareness (H4b).

3.2.5. Information socialization

Information socialization refers to the degree of providing socializing interaction information (Lu & Chen, 2021). Social interaction is able to enhance viewers’ perception of the warmth of a shopping medium and is considered to be an important driver of consumers’ using intention towards a shopping medium and satisfaction with E-services (Hou et al., 2019; K. Kang et al., 2020; Ma, 2021a). Live streaming commerce has prominent functional advantages to facilitate viewers to communicate and socialize with streamers and other viewers in a real-time manner, such as clicking the “like” or “share” buttons, sending positive comments or greetings to streamers or other viewers through bullet screens (Gao et al., 2021). The real-time two way social interaction releases a signal of warmth and gives viewers a feeling of face-to-face communication (Ma, 2021b). The previous studies show that the real-time interaction can shorten the psychological distance between viewers and streamers, and establish intimate relationships among them, thus leading to a high level of trust as well as purchasing intention (Bapna et al., 2017; M. Zhang et al., 2022).

The intensive interaction strengthens the social ties between viewers and streamers (Huang & Suo, 2021), and gives viewers a sense of belonging (Ma, 2021b). According to the social learning theory, individual behaviour is affected by observational learning, perception, and interaction (Bandura & Walters, 1977). In the context of Chinese culture, individuals who are in one group always make decisions collectively (M. Yang et al., 2023). People are much influenced by their acquaintances and tend to develop similar values and purchasing behaviour (Qi & Ploeger, 2021). Some studies indicate that social interaction is positively associated with consumers’ green consumption behaviour (M. Yang et al., 2023). Loebnitz and Grunert (2018) illustrate that the detailed information about product benefits promotes consumers to form positive purchasing behaviour towards nutritious food products if they incorporate health motivation as a part of their values. The health-related information shared in social interaction will evoke viewers’ concern and interest in green food’s benefits to health (Xu et al., 2021), and develop similar health values. The altruistic attributes of green food as well as streamers’ appeal of supporting rural revitalization and promoting sustainable social development, will arouse viewers’ ethical and environmental concerns (Vermeir et al., 2020). The environmental awareness increases the importance of environmentally friendly attributes of green food in consumers’ purchase decisions (Loebnitz & Grunert, 2018). Based on the above analysis, the following hypothesis is proposed.

H5.

Information socialization will positively impact viewers’ health consciousness (H5a) and environmental awareness (H5b).

3.2.6. Health consciousness and environmental awareness

Health consciousness refers to an individual’s concern about his or her health and readiness to take health-related behaviours (Rahamat et al., 2022; Wang et al., 2019). The existing literature presents that individuals’ health consciousness is a key predictor of their health behaviours (Yadav & Pathak, 2016). High health consciousness will increase individuals’ purchasing intention towards green products (Wandel & Bugge, 1997), such as purchasing natural and harmless food over conventional food (Parashar et al., 2023). Some researchers investigate organic food market in developing countries and indicate that health consciousness plays a significant role in consumers’ purchasing decisions of organic food (Chu, 2018; Singh & Verma, 2017; Yadav & Pathak, 2016). It is considered to be the basis and the first factor affecting consumers’ purchasing intention towards organic food (Nagaraj, 2021). Consumers with high health consciousness are more likely to shape the positive attitude and purchasing intention towards organic food (Parashar et al., 2023). In the context of live streaming commerce, consumers’ purchasing intention is directly associated with their perceptions of social interaction and utility (Ma, 2021b), as well as the persuasiveness of information (Gao et al., 2021). Live streaming commerce conveys various effective information signals to reduce product fit uncertainty and product quality uncertainty, leading to viewers’ positive purchasing intention (Lu & Chen, 2021). Meantime, the health-related information about green food will develop and enhance viewers’ concern about health issues, which drives favourable purchasing intention and consumption behaviour towards green food. This implies that health consciousness not only directly leads to the purchasing intention, but also mediates the relationships between information features of live streaming commerce and purchasing intention. Based on the above analysis, the following hypotheses are proposed.

H6.

Health consciousness will positively impact viewers’ purchasing intention.

H7.

Health consciousness will mediate the relationships between information features and purchasing intention.

Environmental awareness refers to an individual’s concern about environmental problems and readiness to engage in environmentally friendly behaviours (Agrawal et al., 2023; Xu et al., 2020). People’s environmental awareness impacts their green behaviours (Sivapalan et al., 2021), and it is able to increase consumers’ purchasing intention towards green products (Paul et al., 2016). The environmental attitude is regarded as an essential factor influencing consumers’ attitude and purchasing behaviour towards green food (Larson, 2018; Tong et al., 2020). The increase in demand for organic and green food is related to the rising environmental awareness (Ahmed et al., 2021; Wang et al., 2019), and people’s altruistic motives drive organic and green food consumption (Paul et al., 2016). A beneficial effect on the environment is one of the essential drivers of organic food consumption (Lăzăroiu et al., 2019). That is, consumers who have a high cognition level of organic food’s benefits to the environment are more likely to have a moral attitude and higher purchasing intention (Nagaraj, 2021). The quality of information, such as richness, completeness, accuracy, reliability, etc., has a strong association with viewers’ purchasing intention (Gao et al., 2021). Moreover, the information and knowledge of green food’s benefits to the environment that viewers obtain from live streaming services, will develop and enhance viewers’ environmental awareness, which in turn increases purchasing intention (Taufique et al., 2016). That is, environmental awareness not only directly leads to the purchasing intention, but also mediates the relationships between information features of live streaming commerce and purchasing intention. Based on the above analysis, the following hypotheses are proposed.

H8.

Environmental awareness will positively impact viewers’ purchasing intention.

H9.

Environmental awareness will mediate the relationships between information features and purchasing intention.

3.2.7. Purchasing intention and purchasing behavior

Purchasing intention is the result of external stimuli and cognitive evaluations and is the predecessor of purchasing behaviour (Parashar et al., 2023). According to the theory of planned behaviour (TPB), an individual’s specific purchasing behaviour is determined by his or her intention towards the products (Rahamat et al., 2022). The studies related to food choices reveal that consumers’ purchasing intention towards healthy and safe food products is the prerequisite of actual purchasing behaviour (Chen et al., 2018; H. V. Nguyen et al., 2021; Wee et al., 2014). Consumers are more likely to purchase green food products if they have a positive purchasing intention. In the context of live streaming commerce, viewers’ purchasing intention will be driven with the increase in health consciousness and environmental awareness. Some researchers present that attitude plays an important role in forming purchasing intention and behaviour towards green and organic food (Ahmed et al., 2021; Qi & Ploeger, 2021; Wee et al., 2014). While, the relationship between attitude, behavioural intention, and actual behaviour might be incongruous (Lee & Yun, 2015; Qi & Ploeger, 2021), which means that consumers might not purchase green food due to product quality uncertainty, even if they have a favourable attitude towards green food. Nevertheless, a positive attitude will enhance the impact of purchasing intention on actual purchasing behaviour. Based on the above analysis, the following hypotheses are proposed.

H10.

Purchasing intention will positively impact viewers’ purchasing behaviour.

H11.

Attitude will strengthen the relationship between purchasing intention and purchasing behaviour.

One main reason that consumers prefer to live streaming shopping is that they are attracted by famous and attractive streamers (Hu et al., 2017). Streamers are considered as “opinion leaders” to convey detailed and effective information to facilitate viewers’ purchase decisions (Ma, 2021a; M. Zhang et al., 2022). Although some researchers indicate that streamer expertise is not directly associated with consumers’ purchasing intention and behaviour (Gao et al., 2021; Park and Lin, 2020), it is considered as a crucial factor influencing live streaming commerce success (Heo et al., 2020) and consumer loyal in the long run, it even leads viewers to become dependent on live streaming shopping. Viewers tend to trust streamers and generate a positive purchasing intention when they think streamers are professional in introducing, evaluating, and recommending products (Lu & Chen, 2021; M. Zhang et al., 2022). The information obtained from expert sources is highly persuasive and is typically perceived as useful to result in a favourable cognitive evaluation (Ma, 2021a). Streamer expertise is particularly vital for consumers’ green food choices if they do not have a high level of green food involvement. Professional streamers are able to promote viewers’ purchasing intention to convert into purchasing behaviour. Based on the above analysis, the following hypothesis is proposed.

H12:

Streamer expertise will strengthen the relationship between purchasing intention and purchasing behaviour.

4. Methodology

4.1. Measures

Construct measurement items were developed referring to the previous related research, with adaptions for the features of green food products. To establish content validity, we invited five professors from the fields of E-commerce and consumer behaviour research, and five experienced live streaming shoppers to review the representativeness and suitability of the items. Then, we modified some statements and expressions according to the analysis of expert evaluation.

Items for information richness were adapted from Sun et al. (2019) and Busalim et al. (2021). Items for information authenticity were adapted from Ma (2021a) and Busalim et al. (2021). Items for information synchronicity and information personalization were adapted from Y. Li et al. (2021) and M. Zhang et al. (2022). Items for information socialization were adapted from Ma (2021b). Items for health consciousness were adapted from Rahamat et al. (2022), and items for environmental awareness were adapted from Xu et al. (2020). Items for purchasing intention and purchasing behaviour were adapted from Wee et al. (2014). Items for attitude were adapted from Singh and Verma (2017). Items for streamer expertise were adapted from Ma (2021a). Each item was rated on a 5-point Likert scale ranging from “extremely disagree” to “extremely agree”.

4.2. Sample and data

For the empirical study, an online survey was conducted from May to June 2022. Since our survey was aimed at investigating consumers’ purchasing behaviour towards green food products, one qualifier question was set: “Have streamers ever recommended green food products when you use live streaming shopping?” Only respondents who answered “yes” were qualified.

An offline pre-tested survey was first conducted with 20 respondents who purchased green food products in live streaming channels over the past month, with the aim of checking the comprehension and logic of the contents of our questionnaire. For enhancing the effectiveness, we gave out some small gifts, such as the shopping bags, coupons, etc. Then, we rewrote the questionnaire to ensure that each survey item was clear and understandable. The formal data collection was conducted by one professional survey company in China, and the coupons (e.g.“Tmall online supermarket”shopping coupons) were used to recruit participants and enhance the completeness and availability of responses. Finally, 948 valid responses were received. Table 2 shows the demographic information of the respondents.

Table 2. Demographics information (N = 948).

Demographics	Categories	Frequencies	Percentages
Gender	Male	455	48.0%
	Female	493	52.0%
Age	≤25	280	29.5%
	26–35	239	25.1%
	36–45	198	20.9%
	46–55	167	17.6%
	56≤	64	6.8%
Education	≤High school	132	13.9%
	Junior College	348	36.7%
	Bachelor	370	39.6%
	Master≤	93	9.8%
Monthly income (Chinese yuan)	≤5000	147	15.5%
	5001–9999	295	31.1%
	10000–14999	183	19.3%
	15000–19999	203	21.4%
	20000≤	120	12.7%
Online shopping experience	≤1 year	228	24.1%
	1–2 years	242	25.5%
	2–3 years	238	25.1%
	3–4 years	147	15.5%
	4 years≤	93	9.8%
Watching frequency for E-commerce ive streaming	Almost none	54	5.7%
	1–3 (per month)	278	29.3%
	1–2 (per week)	290	30.6%
	3–6 (per week)	237	25.0%
	1≤ (per day)	89	9.4%

Of all the respondents, females account for 52%, while males account for 48%. More than half of the respondents are aged below 35 (54.6%), and more than three-quarters have a bachelor degree or associate degree (2- or 3-year degree). The most common household monthly income category is 5,001–9,999 yuan (31.1%). In terms of watching frequency, 29.3% of the respondents watch live streaming videos 1–3 times a month, 30.6% of the respondents watch 1–2 times a week, and 25% of the respondents watch 3–6 times a week. Live streaming shopping is becoming one prevalent shopping pattern, especially for young people with a better education background. These results are similar to the users’ characteristics of live streaming commerce in China (Gao et al., 2021; M. Zhang et al., 2022).

5. Results and discussion

5.1. The measurement model

We executed our model in two steps: the first step was the testing of the measurement for the reliability and validity of the constructs, and the second step was the estimation of the structural model for the proposed hypotheses. A confirmatory factor analysis (CFA) with 11 latent variables and 42 observed variables was conducted. The loading coefficient of an item should be greater than the threshold value of 0.5 (Fornell & Larcker, 1981), so we deleted the items “IR5 streamer provides novel information of green food”, “IA4 Streamer provides authoritative information of green food”, and “IP4 streamer provides personal food customization based on my requirements”. The final model fitness is robust (x² = 1059.946. p < 0.001, x²/df = 1.464; CFI = 0.982; TLI = 0.979; RMSEA = 0.022; SRMR = 0.027), hence the measure model fits the data well. Reliability is measured with the values of Cronbach’s alpha and composite reliability (CR). The threshold value of 0.70 for Cronbach’s alpha suggested by Hair and Alamer (2022) and the value of 0.70 for CR suggested by Fornell and Larcker (1981) are used to assess the reliability of a measurement model. As shown in Table 3, all constructs show good internal consistency reliability with Cronbach’s alpha values ranging from 0.806 to 0.915 and CR values ranging from 0.807 to 0.916.

Table 3. Confirmatory factor analysis.

Construct	Items	Stan. Esti.	S.E.	Cronbach alpha (α)	CR	AVE
IR	IR1	0.738***	0.019	0.823	0.824	0.540
	IR2	0.656***	0.022
	IR3	0.757***	0.018
	IR4	0.782***	0.017
IA	IA1	0.764***	0.017	0.806	0.807	0.584
	IA2	0.824***	0.014
	IA3	0.699***	0.019
IS	IS1	0.702***	0.020	0.829	0.828	0.548
	IS2	0.729***	0.019
	IS3	0.731***	0.019
	IS4	0.795***	0.017
IP	IP1	0.765***	0.017	0.810	0.810	0.587
	IP2	0.739***	0.018
	IP3	0.793***	0.016
SI	SI1	0.686***	0.021	0.830	0.831	0.552
	SI2	0.797***	0.017
	SI3	0.776***	0.018
	SI4	0.707***	0.020
HC	HC1	0.732***	0.019	0.827	0.828	0.547
	HC2	0.779***	0.017
	HC3	0.763***	0.018
	HC4	0.681***	0.021
EA	EA1	0.763***	0.017	0.842	0.842	0.573
	EA2	0.694***	0.020
	EA3	0.788***	0.017
	EA4	0.778***	0.017
PI	PI1	0.733***	0.019	0.830	0.831	0.552
	PI2	0.806***	0.017
	PI3	0.700***	0.020
	PI4	0.728***	0.019
PB	PB1	0.738***	0.020	0.818	0.818	0.601
	PB2	0.772***	0.018
	PB3	0.813***	0.017
SE	SE1	0.871***	0.011	0.897	0.899	0.747
	SE2	0.903***	0.010
	SE3	0.817***	0.013
AT	AA1	0.891***	0.010	0.915	0.916	0.784
	AA2	0.921***	0.009
	AA3	0.842	0.011

Note: *** indicates p<0.001^,**indicates p<0.01, * indicates p<0.05, ns = insignificant.

Convergent validity and discriminant validity are used to assess construct validity of all variables. Convergent validity is analysed by assessing item loadings and the average variance extracted (AVE) for each construct. As shown in Table 3, all of the estimated factor loadings exceed the required threshold value of 0.5 (Fornell & Larcker, 1981) and show significant p-values (item loadings ranging from 0.656 to 0.921, p < 0.001). AVE values of all the constructs are between 0.540 and 0.784 exceeding the required threshold value of 0.50 (Fornell & Larcker, 1981), confirming satisfactory convergent validity. Discriminant validity is assessed by the AVE for each construct being higher than the squared correlation coefficient between two constructs (Fornell & Larcker, 1981), where the results of Table 4 show that any correlation value between two constructs is not higher than the square root of the construct’s AVE. Therefore, all the constructs are considerably different from one another.

Table 4. Discriminant validity.

	IR	IA	IS	IP	SI	HC	EA	PI	PB	SE	AT
IR	0.735
IA	0.448	0.764
IS	0.406	0.334	0.740
IP	0.366	0.422	0.332	0.766
SI	0.353	0.392	0.393	0.365	0.743
HC	0.487	0.385	0.410	0.349	0.377	0.740
EA	0.392	0.414	0.337	0.320	0.416	0.390	0.757
PI	0.351	0.296	0.353	0.279	0.371	0.427	0.321	0.743
PB	0.408	0.360	0.261	0.342	0.399	0.398	0.398	0.322	0.775
SE	−0.024	−0.036	−0.014	−0.009	−0.047	−0.012	−0.034	0.027	−0.010	0.864
AT	−0.010	0.009	−0.067	0.026	0.006	−0.020	0.037	0.012	0.035	0.121	0.885

Note: The diagonal line of the correlation matrix is the square root of AVE.

Marker variable method is used to measure the extent of common method bias. The results in Table 5 show that the values of CFI and TLI decrease, and the values of RMSEA and SRMR increase after marker variable, thus common method bias can be regarded as nonexistent. In addition, the presence of multicollinearity is estimated through the variance inflation factor (VIF). All the VIF values are below 3.3 (Cenfetelli & Bassellier, 2009), indicating that the multicollinearity problem is nonexistent.

Table 5. Common method bias analysis.

Model fitness	x^2	p	x^2/df	CFI	TLI	RMSEA	SRMA
Basic model	1059.946	P<0.001	1.464	0.982	0.979	0.022	0.027
Marker variable	2004.393	P<0.001	3.673	0.931	0.925	0.042	0.120

5.2. The structure model

The results show that the structural model is a good fit: x² = 923.266. p < 0.001, x²/df = 1.719; CFI = 0.973; TLI = 0.970; RMSEA = 0.028; SRMR = 0.064, and most of the path weights between the constructs are statistically significant (Table 6 column 2). Information richness has significantly positive effects on health consciousness and environmental awareness (β = 0.293, β = 0.171, p < 0.001, respectively); information authenticity has significantly positive effects on health consciousness and environmental awareness (β = 0.107, p < 0.01; β = 0.191, p < 0.001, respectively); information synchronicity has significantly positive effects on health consciousness and environmental awareness (β = 0.181, p < 0.001; β = 0.102, p < 0.05, respectively); information personalization has a significantly positive effect on health consciousness (β = 0.093, p < 0.05); and information sociability has significantly positive effects on health consciousness and environmental awareness (β = 0.144, β = 0.225, p < 0.001, respectively), thereby supporting H1, H2, H3, H4a and H5. Information personalization does not significantly impact environmental awareness, H4b is not supported.

Table 6. Results of the structural model.

Relationship	Model 1	Model 2	Model 3	Model 4
IR→HC	0.284***	0.284***	0.284***	0.284***
IA→HC	0.093*	0.093*	0.093*	0.093*
IS→HC	0.178***	0.178***	0.179***	0.179***
IP→HC	0.083*	0.083	0.082	0.082
SI→HC	0.142***	0.142**	0.142**	0.142**
IR→EA	0.178***	0.178**	0.178**	0.179**
IA→EA	0.180***	0.180***	0.180***	0.180**
IS→EA	0.108*	0.108	0.107	0.107
IP→EA	0.063	0.063	0.064	0.064
SI→EA	0.239***	0.239***	0.240***	0.240***
HC→PI	0.431***	0.433***	0.432***	0.434***
EA→PI	0.221***	0.221***	0.223***	0.223***
PI→PB	0.334***	0.344***	0.354***	0.362***
SE→PB	—	−0.016	—	−0.018
PI*SE→PB	—	0.134**	—	0.112*
AT→PB		—	0.021	0.023
PI*AT→PB		—	0.186***	0.176***

Notes: ***p < 0.001; **p<0.01; * p<0.05; ns = insignificant.

Results also indicate that health consciousness and environmental awareness have positive effects on purchasing intention (β = 0.390, β = 0.215, p < 0.001, respectively); purchasing intention has a positive effect on purchasing behaviour (β = 0.359, p < 0.001), thereby supporting H6, H8 and H10. Our results verify that both health consciousness and environmental awareness are crucial drivers of purchasing intention and behaviour towards green food.

5.3. Mediating effects

To estimate whether information features of live streaming commerce impact purchasing intention through health consciousness and environmental awareness, bootstrapping procedures in Mplus7.4 are employed to determine the mediating effects of health consciousness and environmental awareness. The results (Table 7 column 4) show that health consciousness and environmental awareness together mediate the impacts of information features (except for personalization) on purchasing intention. Although we state that the two mediating variables do not significantly mediate the relationship between information personality and purchasing intention, the regression results show that the mediating effects are at a 10% significant level. The specific indirect effects (Table 7 column 3) of environmental awareness are not significant, therefore H9 is not supported. Information richness, synchronicity, and sociability indirectly impact viewers’ purchasing intention through health consciousness, therefore H7 is partially supported.

Table 7. Results of the mediated path analysis.

Path	Direct	Specific indirect	Total indirect	Total
IR→HC→PI	0.086	0.066**	0.078**	0.164**
IR→EA→PI		0.012
IA→HC→PI	0.026	0.024	0.038*	0.064
IA→EA→PI		0.014
IS→HC→PI	0.120*	0.038*	0.044*	0.165**
IS→EA→PI		0.007
IP→HC→PI	0.041	0.021	0.026	0.067
IP→EA→PI		0.005
SI→HC→PI	0.156*	0.031*	0.047*	0.204***
SI→EA→PI		0.016

Note: x²=761.716, p<0.001, x²/df =1.437; CFI=0.984; TLI=0.982; RMSEA=0.021; SRMR=0.033.

***indicates p<0.001^,**indicates p<0.01, * indicates p<0.05, ns = insignificant.

5.4. Moderating effects

The results of SEM with moderators are presented in Table 6. The columns 3–5 show the results of employing streamer expertise as the moderating variable, employing attitude as the moderating variable, and employing both streamer expertise and attitude as the moderating variables, respectively. The coefficients of interactive items in columns 3–5 are positive and statistically significant. That is, both consumer attitude and streamer expertise favourably moderate the impact of purchasing intention on purchasing behaviour. The results of column 5 show that attitude and streamer expertise significantly strengthen the relationship between purchasing intention and purchasing behaviour towards green food (β = 0.176, p < 0.001; β = 0.112, p < 0.05, respectively), thereby H11 and H12 are supported.

6. Discussion

Information richness, authenticity, synchronicity, and sociability significantly promote the formation of health consciousness and environmental awareness (H1, H2, H3 and H5). The positive impacts of live streaming features on viewers’ psychological responses (e.g. cognitive attitude, trust, satisfaction, attachment, etc.) have been proven in the previous studies (Gao et al., 2021; Y. Li et al., 2021; Ma, 2021a; M. Zhang et al., 2022). Qi and Ploeger (2021) suggest that improving accessibility of useful and objective information, emphasizing the nutritious and ecological benefits of green food products, and providing tactile and visual cues to consumers will drive favourable purchasing intention. The more product knowledge consumers have, the more likely they are to make environmentally friendly purchasing decisions (Cachero-Martínez, 2020; R. Yang et al., 2020). As consumers obtain more organic and green food information, they are more concerned for their health and generate purchasing intention (Ayyub et al., 2021; Parashar et al., 2023). Thus, there lies a strong association between the information that viewers obtain, health consciousness, environmental awareness, and purchasing intention that leads to purchasing behaviour. Our findings show that the abundant and authentic information, synchronic interaction, and social communication improve viewers’ cognition of green food’s benefits, increase their knowledge about health and environmental issues, and drive higher health consciousness and environmental awareness. Our findings further verify the influences of different types of the information on consumers’ psychological responses. In particular, information richness exerts a stronger impact on health consciousness, which is probably because the complete information is more persuasive (Gao et al., 2021), making viewers believe that green food is appropriate to bolster health, and arousing viewers’ concern about health issues. Information sociability plays a significant role in forming environmental awareness, which is in line with the previous study on social interaction’s positive impact on environmental awareness as well as green practices (Agrawal et al., 2023). Chinese consumers who are in one group are more likely to influence each other (Qi & Ploeger, 2021). The intimate social interaction in live streaming channels has prominent influences on developing environmental awareness with the increasing public concern on environmental issues. An insignificant relationship is found between information personalization and environmental awareness, which is probably because personalized communication focuses more on the topics related to viewers’ targeted products and own needs, so it significantly promotes the formation of health consciousness.

Both health consciousness and environmental awareness positively impact viewers’ purchasing intention towards green food (H6 and H8). These results are similar to the previous research, which indicates that health consciousness and environmental awareness significantly improve consumers’ intention to purchase green products, such as green furniture (Xu et al., 2020) and organic food (Ayyub et al., 2021; Parashar et al., 2023). Some researchers illustrate that Chinese consumers’ green food buying intention is more closely related to the utilitarian evaluations, such as their perceptions of nutritious, natural and safe contents (Qi & Ploeger, 2021), which is because people who purchase green or organic food are mainly motivated by food’s health-related attributes (Chu, 2018; Dangelico et al., 2021). Health consciousness is considered as a crucial driver of consumers’ purchasing intention towards organic food (D. T. Nguyen & Truong Dt, 2021). Our results also show that viewers’ purchasing intention is highly associated with health consciousness, while the importance of environmental awareness’s impact on purchasing intention should be emphasized with the increasing concern of environmental issues (Parashar et al., 2020; Tong et al., 2020). Consumers with high health consciousness pay more attention to food safety and want food products to be produced and processed in an environmentally friendly manner (Loebnitz & Grunert, 2018; Tong et al., 2020). These consumers are more likely to care about environmental issues and possess high environmental awareness promoting to purchase environmentally friendly products (D. T. Nguyen & Truong Dt, 2021; Sadiq et al., 2021). Our results show a significantly positive relationship between purchasing intention and purchasing behaviour (H10), which is similar to the previous research related to organic food purchase (Parashar et al., 2023; Wee et al., 2014).

The mediating effects of health consciousness and environmental awareness are validated in this study. The results reveal that the various information of green food’s benefits to health and the environment drives viewers’ concerns of health and environmental issues, which promote the formation of purchasing intention and behaviour. The formation of purchasing intention is the result of the combined impacts of health consciousness and environmental awareness, which is consistent with the previous research about organic food purchasing behaviour (Parashar et al., 2023). Further analysis reveals that the specific mediating effect of health consciousness is found to be stronger. The relationships between information richness, synchronicity, sociability and purchasing intention are totally or partially mediated by health consciousness. The richer, synchronic and social information provided by live streaming services significantly enhances viewers’ concern about health issues, and then promotes the formation of purchasing intention.

This study also validates that streamer expertise and consumer attitude moderate the relationship between purchasing intention and purchasing behaviour. Our statistical analysis shows that consumer attitude strengthens the impact of purchasing intention on purchasing behaviour. The previous research proves the positive association between attitude, purchasing intention and purchasing behaviour towards organic food (Chu, 2018; Nagaraj, 2021). Our study corroborates with the previous research that a favourable attitude contributes to generating actual purchasing behaviour (H. V. Nguyen et al., 2021). Furthermore, our results also highlight that streamer expertise strengthens the impact of purchasing intention on purchasing behaviour. Viewers are inclined to believe professional streamers and form positive purchasing intention (Lu & Chen, 2021; M. Zhang et al., 2022). The statistical analysis shows that the coefficient of interactive item (SE*PI) is positive and significant, which means that professional streamers are able to further promote viewers who intend to purchase green food to become real buyers. The coefficient of streamer expertise is negative but not significant, which is probably because streamers are not considered as experts in the field of green food in most situations. This phenomenon does not have a significantly negative impact on purchasing behaviour towards green food due to the prominent information advantages of live streaming commerce. Nevertheless, it is worth mentioning that the importance of streamer expertise need to be paid more attention to in the long run.

7. Conclusions

7.1. Theoretical implications

Live streaming marketing is regarded as a successful way to promote products sales in China. This study focuses on a special product category, green food, assesses the feasibility and advantages of live streaming commerce as a medium to sell green food products, and conducts an empirical analysis to explore consumers’ purchasing intention and behaviour towards green food. A theoretical framework is proposed by extracting five variables from live streaming commerce characteristics and employing health consciousness and environmental awareness as the antecedents of purchasing intention and behaviour. It not only explores the impacts of information features on viewers’ health consciousness, environmental awareness, as well as purchasing intention and behaviour, but also incorporates consumer attitude and streamer expertise as moderating variables to assess their influences on the relationship between purchasing intention and purchasing behaviour. The conclusions will be theoretically conducive to consumer psychology and behaviour research with respect to green food in the context of new commerce pattern and marketing strategy.

This study estimates the influencing mechanism of information features of live streaming commerce on purchasing intention and behaviour towards green food by using health consciousness and environmental awareness as mediating variables, which can help us get more practical and targeted conclusions. Further, this study estimates the moderating effects of streamer expertise and consumer attitude, and verifies that the two variables are vital parameters influencing consumers’ green food choices when they use live streaming shopping. The novel model expands the empirical research of green food consumption from the perspectives of specific marketing patterns, sellers and consumers characteristics, which provides more research directions for the future study.

7.2. Managerial contributions

Our findings will help streamers and green food marketers improve marketing strategies and design more pertinent live streaming contents and activities according to products characteristics. Based on the results that information features of live streaming commerce exert positive influences on health consciousness and environmental awareness, and health consciousness and environmental awareness significantly increase viewers’ purchasing intention and behaviour towards green food products. Streamers and marketers should capitalize on these findings and strive to guide viewers to develop and enhance their concerns about health and environmental issues. Streamers and green food marketers should make full use of live streaming advantages to convey rich, comprehensive, authentic, and credible health and environment-related information to viewers, and emphasize the benefits of green food consumption to social well-being and sustainable development to help viewers shape a positive attitude and purchasing intention. Streamers and marketers should design participatory activities to strengthen viewers involvement and interaction, encourage viewers to express their comments and needs, and share their opinions and experiences through bullet screens to enhance information usefulness and effectiveness. Meanwhile, streamers should provide positive responses and feedback to viewers to establish close emotional links and drive potential needs of green food. In addition, steamers should select products prudently, build partnerships with trusted green food suppliers to provide high quality green food products to viewers.

Our findings provide important implications for streamers and marketers to optimize marketing strategies based on product categories, and streamers and consumers characteristics. As there lies a gap between consumer interest and purchasing behaviour towards green food, the finding of the positive moderating effect of streamer expertise between intention and behaviour will help green food suppliers make effective marketing strategies through training and fostering professional streamers in the long run. Beyond that, green food suppliers and sellers should consider consumers’ shopping traits, and focus on live streaming contents and styles that are associated with health and environment to find potential consumers who might have positive attitudes towards purchasing green food.

The results will help the relevant authorities to make effective policies to increase green farming practices. Such policies help people enhance health consciousness and environmental concern through propagandizing green food’s benefits with respect to health improvement and social sustainability. In addition, some policies should be developed to improve the enthusiasm and business practices of live streaming platforms and streamers towards green food, which in turn further benefit multiple social practices on a large scope through their efforts.

7.3. Limitations and further research perspectives

Some limitations may exist in this study. First, this study employs streamer expertise and consumer attitude as moderating variables in our model, while other potential moderating variables, such as consumers’ demographic characteristics, individuals’ consumption experiences, etc., could be examined in the future studies. Individuals’ health consciousness and environmental awareness might be influenced by these variables, future research in this direction will give more into how their relationships vary in different situations. Second, some other features of live streaming commerce, such as the entertainment of live streaming services, the attractiveness of price privileges, etc., might have important impacts on purchasing intention and behaviour towards green food. Compared with traditional online sellers, streamers are able to get better financial privileges to share with viewers due to their huge followers. Future research included these variables will provide more significant implications. Third, this study explores the determinants of purchasing intention and behaviour towards green food from the consumer perspective, future research that considers the insights of agricultural companies, government, and other stakeholders will provide a broader research framework for the development of green food industry in emerging economies.

Abbreviations

IR	=	Information richness
IA	=	Information authenticity
IS	=	Information synchronicity
IP	=	Information personalization
SI	=	Information socialization
HC	=	Health consciousness
EA	=	Environmental awareness
PI	=	Purchasing intention
PB	=	Purchasing behaviour
SE	=	Streamer expertise
AT	=	Attitude

Acknowledgments

This work was supported by The Philosophy and Social Science Foundation of Shandong Province, China [Grant No. 21CGLJ34].

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

The work was supported by the Shandong Office of Philosophy and Social Science [21CGLJ34].

Appendix1.

Constructs and Items

Information richness (Sun et al. (2019); Busalim et al. (2021))

IR1 Streamer provides detailed information with respect of green food’s functional merits

IR2 Streamer provides substantial information with respect of green food’s production, process and package, etc.

IR3 Streamer provides vivid sensory experience information of green food

IR4 Streamer provides dynamic demonstration information of green food in may ways

Information authenticity (Ma (2021a); Busalim et al. (2021))

IA1 Streamer provides authentic information of green food

IA2 Streamer provides accurate information of green food

IA3 Streamer provides credible information of green food

Information synchronicity (Y. Li et al., 2021; M. Zhang et al., 2022)

IS1 Live streaming platform processes my comments inputs very quickly.

IS2 I feel that seeing comments sent by other viewers is very fast

IS3 I am able to obtain the information I want without any delay

IS4 Streamer responds to my comments inputs very quickly.

Information personalization (M. Zhang et al., 2022)

IP1 Streamer provides the information of all alternative products I intend to buy

IP2 Streamer can help me establish my green food needs.

IP3 Streamer provides relevant information based on my concerns and interests

Information socialization (Ma, 2021b)

SI1 Live-stream platform provides me with opportunities to communicate with streamers and sellers

SI2 Live-stream platform provides me with opportunities to communicate with other viewers

SI3 Social interaction information enables me to feel closer to streamer

SI4 Social interaction information enables me to feel part of the community

Health consciousness (Rahamat et al., 2022)

HC1 I am someone who is concerned about the long-term health effects of my food choices

HC2 I am someone who carefully considers health consequences of my food choices

HC3 I am a health conscious person

HC4 I reflect about my health a lot

Environmental awareness (Xu et al., 2020)

EA1 I am worried about the world’s environment and its consequences for my future

EA2 The balance of nature is very delicate and easily upset

EA3 I am willing to control my consumption to realize the sustainable development

EA4 I think individuals have the responsibility to protect the environment

Purchasing intention (Wee et al., 2014)

PI1 I would buy green food products in the near future

PI2 I plan to buy green food products in regular basics

PI3 I intend to buy green food products for my long term health benefits

PI4 I intend to buy green food products because they are more environmentally friendly

Purchasing behaviour (Wee et al., 2014)

PB1 I often buy green food products

PB2 I often buy green food products for my health

PB3 I often buy green food products because they are more environmentally friendly

Streamer expertise (Ma, 2021a)

SE1 Streamer is very knowledgeable about green food brands and products

SE2 Streamer is an expert on green food products or services

SE3 I feel that streamer is competent enough to make assertions about things that they are

good at

Attitude (Singh & Verma, 2017)

AT1 I think green food is very useful to meet my needs

AT2 I think green food products have higher quality than conventional ones

AT3 I am convinced the consumption of green food is a reasonable action