Research on the influence mechanism of anchors’ professionalism on consumers’ impulse buying intention in the livestream shopping scenario

ABSTRACT For the emerging marketing mode of livestream shopping, this study constructed a model to empirically analyze how anchors’ professionalism affects consumers' impulse buying intention. The results have shown that professional level of anchors is positively correlated with the impulse buying intention of consumers and a sense of presence partially plays an intermediary role in the interaction between anchors’ professionalism and consumers’ impulse buying intention. In addition, consumers’ competitive arousal has been found to have a negative moderating effect on the relationship between anchors’ professionalism and consumers’ sense of presence.


Introduction
The encounter of web celebrities with livestreaming broadcast and e-commerce has produced quite a dazzling spectacle of consumption. As reported in CNNIC's 44th Statistical Report on China's Internet Development, livestreaming e-commerce has become a new bright spot in the growth of online consumption and changed the operation strategy and business model of the whole market. Many groups, such as superstars, corporate executives, and government officials, have started to sell products in the form of livestreaming e-commerce. However, a superficial boom doesn't mean real prosperity. Many superstars with tens of millions of followers failed to drive consumers' purchase intention through livestreaming e-commerce. The possible reason for this phenomenon is that the professional levels of anchors, which greatly affect the process of stimulating consumers' impulse buying intention, are uneven in the scenario of livestreaming e-commerce.
After reviewing the research findings of consumers' impulse buying intention under the scenario of livestreaming e-commerce, it has been found that existing researches mainly focus on exploring the influence of the external stimuli for consumers' impulse buying intention, such as the convenience of mobile payment (Liao and Ho 2021), the website brand (Chang and Chen 2008), consumers' personalities (Liu, Li, and Hu 2013), and online shopping atmosphere (Liu, Li, and Yin 2018). Following the stimulus-organismresponse research framework, Zhang and Zheng (2019) found that the price, quality and promotion of commodity are both influential roles in consumer purchase behaviour (Zhang and Zheng 2019). Li et al. (2016) revealed that the atmosphere in virtual communities, including its informativeness and entertainment, have a positive impact on online impulse buying intention (Li, Cui, and Cheng 2016). However, according to questionnaire suvey, Yang (2021) argued that product display makes an influence of consumers' choice but most consumers are not satisfied with the products purchased through the way of live video streaming (Yang 2021). Besides, both of real-time promotions and anchor opinion leader traits have also been proven to be very important to users' purchase intentions and behaviours (Zhou, Lei, and Zou 2011). But we are not very clear about why the characteristics of the anchor are to affect the behaviour of consumers, and which characteristics of the anchor will affect the behaviour of consumers. This is exactly what this paper discuss.
In the scenario of livestreaming e-commerce, all commodity transactions are the result of the joint action of humans, goods, and environments, manufacturers produce and store goods, businesses display goods on the platform, anchors sign contracts with employment agencies (MCN) for commodity promotion, and users purchase goods through multiple channels. From this perspective, anchors play a key role that can not be ignored in the scenario of livestreaming. The selection, processing, production, and transmission of information by anchors all affect consumers' attention, understanding, and impulse buying intention (Chen and Zhang 2019). For consumers, they are eager to get professional information and knowledge from experienced people. But sometimes, under the temptation of economic interests, some anchors tend to attract consumers' attention by making jokes, hyping new products, selling fake products, and so on to gain huge profits. As a result, the negative image of anchors such as a lack of professionalism has been criticised (Chen and Liu 2017). The lack of unified management of anchors' professional qualities will lead to the uneven performance of anchors, which will further restrain consumers' impulse buying intention in the livestreaming e-commerce scenario. In the huge livestreaming e-commerce market, it is difficult to have professional anchors accordingly, which is a big challenge faced by the practitioners in the live streaming e-commerce industry. At present, the academic researches on e-commerce livestreaming mostly focus on the mechanism of action, or on the description of the emerging marketing phenomenon of social media influencer, and there are few relevant empirical researches. Few studies focused on the role of anchors in the livestreaming e-commerce scenario or the relationship between anchors' characteristics and consumers' behaviour. This paper is aimed at exploring the underlying influence mechanism of anchors' professionalism on consumers' impulse buying intention, revealing the important role of professional anchors in the livestreaming e-commerce scenario, and providing a theoretical basis for the practice of more effective management optimisation of livestreaming e-commerce.
Researches on both offline and online shopping have provided a great deal of evidence for the key role of personal affective factors in determining impulse buying intention (Chang and Chen 2008) (Liu, Li, and Hu 2013). AI-Tit AA et al. (2020) suggest that social support and trust are the key driving force of social commerce intention and can be used to motivate social media users to make purchase decisions (Al-Tit, Omri, and Hadj 2020). However, livestreaming marketing is a new way to purchase for consumers, especially after 2020, the impact of COVID-19 has hindered traditional offline marketing. The convenience of mobile payment makes e-commerce more and more important in people's lives. And also, personalised recommendation systems (PRS) use the user's interests and purchasing behaviour, to recommend information and merchandise for users (Liao and Ho 2021). This has gone a long way to making online shopping easier. But With the expansion of e-commerce and the number of goods available online, customers require time to find the goods that they want to buy. Browsing of a large number of irrelevant information sources and product processes means that consumers experience information overload. It is difficult for people to choose the products they really like and need to drink from a complex and diverse range of products. And different from traditional e-commerce, livestreaming e-commerce is an interactive, engaging, usercentric and synchronous environment that offers real-time human computer-mediated interaction between viewer and streamer (e-vendor) as well as viewer-viewer (Hilvert-Bruce et al. 2018)(Clement Addo, Fang, andO et al. 2021). Zinko et al. (2021) supported that images have a greater impact on online consumers' consumer behaviour than text (Zinko, de Burgh-woodman, and Furner et al. 2021). And the impact of live streaming media on the consumers' behaviour of online consumers is likely to be greater compared to that images. Live streaming and social commerce have provided new opportunities and challenges to academia, businesses, and consumers (Felix, Rauschnabel, and Hinsch 2017) (Mcmillan 2013). Live streaming enhances digital marketing credibility because consumers engage in real-time chats with the e-vendors, motivating consumers to pay attention to the focal product. Live streaming also presented an opportunity for small retailers to have a closer connection with consumers. It is important to note that closer engagement with e-vendors is vital in making a purchase decision (Ho and Rajadurai 2020). Previous studies have shown that information transmission through new social media could make people lose the opportunity to have face-to-face emotional communication, reduce the real perception of interpersonal interaction, and thus weaken a sense of presence (Chen 2007). However, from the perspective of the Media Richness Theory, a high level of richness of social media helps users perceive the existence of interactive objects and have a sense of presence in the process of use. The absence of physical distance and the illusion of face-to-face communication are also favourable for shortening the psychological distance between users (Dai and Liu 2015). There is still no unified understanding of the role of a sense of presence in information transmission in the scenario of livestreaming e-commerce. The existing researches are scattered, most of which only involve traditional online shopping situations, and there are relatively few studies on the sense of presence perceived by consumers in the livestreaming e-commerce scenario or the systematic empirical analysis. Therefore, it is necessary to carry out a more in-depth analysis of the influence of consumers' sense of presence in the situation of livestreaming e-commerce.
Moreover, one of the main ways to increase real-time sales in the scenario of livestreaming e-commerce is that anchors arouse consumers' sense of competition through professional product recommendation skills. When participating in online shopping, many consumers purchase products that simultaneously satisfy their emotional needs of conformity, desire for a counterattack, and prying into privacy (Meng et al. 2020). Competitive pressure in group consumption easily stimulates consumers' herding behaviour (Chien 2016), as well as the desire to win or lose in impulse buying. Despite the important role of a sense of competition acknowledged by some scholars in the research area of consumer behaviour, the phenomenon of many consumers' participation in snapping up products through the real-time network is more incisively and vividly reflected in the scenario of livestreaming e-commerce than in the traditional online shopping context. Consumers can more directly feel the competitive pressure of purchase through the livestreaming e-commerce interface. The existing studies have not comprehensively reached the role of consumers' competitive arousal in the livestreaming e-commerce scenario. And the following questions also need to be explored urgently: Can the professional recommendations of anchors more effectively arouse consumers' sense of competition and then stimulate their impulse buying intention? Can the factor of consumers' competitive arousal enhance consumers' perception of immersion in the livestream shopping room and then strengthen consumers' sense of presence. An in-depth discussion on these issues will not only enrich the theory of consumer decision-making but also provide guidance on the organisation of more effective promotional activities for livestreaming e-commerce.
To sum up, the focus of this study is on anchors' professionalism, a key factor in the scenario of livestreaming e-commerce. The research sample is composed of 374 consumers with livestream shopping experiences. Based on the questionnaire survey, the regression analysis was conducted to empirically analyse the relationship between the professional level of anchors and consumers' impulse buying intention. Presence was introduced as a mediating variable to investigate the influence of anchors' professionalism on consumers' sense of presence. Furthermore, the moderating variable of consumers' competitive arousal was used to explore its moderating effect on the relationship between anchors' professionalism and consumers' sense of presence. As a result, this study has revealed the operating mechanism underlying the livestreaming e-commerce economy and provided some practical guidance and suggestions for the existing practical problems in the field of livestreaming marketing.

Anchors' professionalism and consumers' impulse buying intention
According to Social Psychology, the characteristics of information sources include three aspects: credibility, professionalism, and attractiveness. Sussman and Siegel (2003) have proposed two key factors affecting effective information adaption, including the quality of information and the credibility of the information source (Sussman and Siegal 2003). The professionalism of information providers will directly affect the persuasiveness of information (Kelman 1961). The higher the quality of persuasive information provided by the information system for an individual is, the higher the level of the information's social influence perceived by the individual will be (Cho and Chan 2019). Moreover, Yang et al. (2017) believe that the professional levels of information sources are determined by the rigorous and inviolable information content transmitted (Yang and Shen 2017). Meng et al. (2020) have made a new interpretation of Internet celebrities' professionalism, which refers to the relevant knowledge summarised by Internet celebrities based on their learning of fashion information and extensive use experience (Meng et al. 2020). In the scenario of livestream shopping, the anchor is the main information sender, namely the information source. Thus, the professional level of the anchor greatly affects information transmission and consumers' acceptance of the information. The more professional the anchor's image perceived by consumers is, the higher the professional degree of the anchor believed by consumers will be.
In general, it is difficult for consumers to make an objective evaluation in the face of complex information presentation when they are shopping online (Yang et al. 2018). According to the research by Chen et al. (2019), professional recommendations will directly affect consumer's opinions on products and their purchase behaviours (Chen and Zhang 2019). Professional anchors need to have this kind of ability The anchor usually has certain experience and knowledge accumulation in products and live broadcast techniques. In the scenario of live broadcast, both e-commerce anchor and the audience can become the opinion publishers, but the effects are different due to the different publishing identities of publishers. Anchors have natural advantages in the way of images and sounds, compared with the audience who can only express their opinions by words. Luo et al. (2012) proposed that interpersonal network has asymmetric characteristics (Luo and Xi 2012). And the feature will be intensified in the scenario of livestream shopping which is reflected in the fact that individuals in the network do not have exactly the same right of discourse, while those in the centre of the hole in the network structure have more influence. This centrality acts as an opinion leader, influencing others through interpersonal network. Anchors plays such a role. Therefore, hypothesis 1 is proposed in this paper: H1: In the scenario of livestream shopping, anchors' professionalism positively affects consumers' impulse buying intention.

The mediating role of presence
Presence plays an important role in improving the online shopping experience (Zhang 2010). With the development of virtual reality technology, a sense of presence has been regarded as one of the main characteristics of media. Different websites, technologies, structures, and contents will produce different forms of presence (Zheng 2019). According to the theory of social network, people will have a sense of presence for the surrounding perception when they enter a new environment, which will affect their behaviour. In the scenario of livestream shopping with user interaction function, this feature is strengthened. During online shopping, presence is the key mechanism that stimulates consumers' emotional consumption (Cyr et al. 2007). Research shows that a strong presence can stimulate people's imagination of their interpersonal communication with other people (Lowenthal 2009), as well as their positive attitude towards the content of communication (Kato, Kato, and Akahori 2007). But on the other side, online shopping can inhibit consumers' imagination and thus affect their feelings of pleasure during online shopping.
Studies have concluded that three elements are required to maintain a high level of presence in online shopping: a computer as an intermediary, an appropriate software system as an interface for consumers to obtain information and some interactive channels to transmit information (Meng et al. 2020). Compared to the traditional online shopping via website, the form of livestream shopping not only satisfied the use of computers as the main medium of information dissemination, but also satisfies the interactive behaviour in the process of information transmission, it has a greater impact on people's sense of presence. In e-commerce live broadcast, the anchors of e-commerce mostly encourage users to interact with the anchors in the live broadcast scene by means of speech induction, so as to achieve the goal of increasing the popularity of the broadcast room and users obey the relevant marketing behaviours of the anchors (Khalifa and Shen 2004). When the emotional response of consumers is stimulated, they will be more likely to produce a positive attitude towards transactions (Argo, Dahl, and Manchanda 2005). In particular, in the livestream shopping situation, which is an important environmental cue, a sense of presence may help optimise the virtual shopping experience of consumers in the process of shopping, enhance their perception of authenticity and reliability of the livestream shopping scenario, and thus promote their impulse buying intention.
H2: In the scenario of livestream shopping, presence positively affects consumers' impulse buying intention.
A sense of presence is an important concept to measure consumers' perceived experience in a virtual environment. Presence emphasises that intimacy and directness play an important role in the perception of 'a real person', it has a strong application value in online shopping (Zhao, Wang, and Zhou 2015). The eye contact, smile, friendliness and other interactions that consumers need in real-life scenarios cannot be obtained in traditional online shopping methods, but such feelings can be obtained in livestream shopping. All the participants in a livestream will have a sense of participation, and there will be a potential ritualistic communication between anchors and consumers (Yan 2019). Since the presence of anchors replaces the physical presence of consumers, and thus anchors can control the emotions of consumers through their professional skills and make them switch between the real world and the virtual world presented by the network, so as to strengthen the sense of presence generated by the combination of reality and virtual reality. Compared with product introduction information from the 'cold' website pages, anchors' real feelings of introduction and explanation obtained by consumers from the 'cold' website pages, anchors' real feelings of the product and their interactions with consumers can provide consumers with a better shopping experience. Through their observations on multiple livestreaming e-commerce platforms, professional anchors with a high level of professionalism can provide consumers with more abundant and professional commodity information, and they know how to effectively interact with consumers on the other end of the video, so they can arouse their interest in the product, reducing their uncertainty in the virtual space, overcoming the virtual network distance, enhancing their sense of presence perceived by consumers in the live broadcast space, and making them feel as if they were interacting and talking with salesmen in a real shopping mall. Anchors can convey their user experience to consumers more directly, become the key opinion leaders, and guide consumers to purchase products. To be specific, firstly, professional anchors usually spend a lot of time trying experiencing products from the standpoint of consumers and acting as product experience officers of consumers, which can not only facilitate consumers' emotional identification but also enhance anchors' ability to evoke consumers' presence. Secondly, professional anchors usually can better understand the consumer needs and have a better knowledge of interaction skills in livestreaming marketing. Although it is not a real face-to-face communication, anchors can create a safe psychological environment for consumers in the situation of online shopping through live barrage interactions, so as to attract consumers to get involved in the livestreaming environment. The stronger the sense of presence perceived by consumers, the more joy they will generate (Barfield, Zeltzer, and Sheridan et al. 1995). This kind of joy will make it easier for consumers to develop a strong purchase desire. Thirdly, professional anchors are more likely to become the key opinion leader of consumers. When shopping online, consumers are trapped in a dilemma of choices caused by the vast number of products. Professional anchors will select different categories of economical products through their individual or team experience, trials, and comparisons to recommend them to consumers in their livestreams. When consumers perceive their professionalism, they will be more likely to have positive feedback on the products recommended by anchors. To sum up, anchors with a high level of professional are more capable of arousing consumers' sense of presence in the virtual environment. Such a strong sense of presence will make consumers develop a closer relationship with the anchors and their recommended products, finally generating strong impulse buying intention.
H3: In the scenario of livestream shopping, anchors' professionalism positively affects presence.
H4: In the scenario of livestream shopping, presence plays a mediating role in the interaction between anchors' professionalism and consumers' impulse buying intention.

The moderating effect of consumer competitive arousal
In the scenario of livestreaming e-commerce, the relationship between anchors' professionalism and presence is not invariable and could be moderated by an important variable, consumer competitive arousal. Most of the relevant studies on the relationship between anchors' professionalism and presence have shown that the establishment of the positive promotion hypothesis mainly depends on the positive interaction between anchors and consumers (Meng et al. 2020) (Dai, Yuan, and Fang 2019) (Dholakia and Zhao 2010). However, some studies have demonstrated that when consumers perceive strong competition during their shopping, this sense of competition will affect their psychological state (Valenti 2006) and stimulate them to make irrational decisions (Nichols and Flint 2010). This happens to be similar to a concept proposed by Nichols and Flint in 2010, namely consumer competitive arousal (CCAr). CCAr is a feeling or thought about the competitive nature of purchase situations, as well as a belief that in a given purchase situation, one needs to compete with other buyers to achieve goals. They believed that this feeling of consumers will generally appear in the process of consumption.However, few studies have explored the sense of competition in the emerging shopping scene of livestream shopping.
This study believes that the sense of competition also exists in livestream shopping and it may have a certain impact on consumers' purchase intentions. First of all, the mobile terminal screen rendering the real-time online numbers and the PINs can make consumers directly perceive the existence of the competitors and the scarcity of the product. With an increase in the online population and a decrease in commodity inventory, real-time consumption images make consumers have a strong sense of co-presence competition (Chen and Han 2018), and consumers immersed in the panic of buying may easily ignore anchors' professional level. Secondly, anchors will create an atmosphere of competitive buying by limiting the purchase time or sales to stimulate consumers to regret not buying the product. The strong feeling of competition perceived by consumers makes it difficult for them to make rational judgements on the products recommended by anchors in a short time and will weaken their senses of insecurity and distrust that may be brought by the virtual network space, thus enhancing the degree of their immersion in livestreams. Thirdly, consumers' feeling of competition can be evoked by 'dan mu' ('bullet screens' in Chinese). Successful buyers may send 'dan mu' to express their joy, which turns turning livestream shopping into a win-lose game, and stimulates consumers' desire to win. Therefore, the focus of consumers shifts from anchors to whether they can successfully buy commodities, and the importance of anchors in the livestream is thus weakened. To sum up, compared with traditional online shopping, livestream shopping provides consumers with a stronger sense of competition. When a group of people are snapping up the same product, the target is not the product but the fun of shopping. Arousing competition among consumers requires the understanding of the inner demand of consumers, the creation of a more immersive atmosphere and a stronger desire to win, and the provision of a sense of presence. In this way, consumers will pay more attention to their competition with others and ignore the professional explanation and recommendation of anchors. The influence of anchors' professionalism on consumers' perception of presence is weakened accordingly.
H5: In the scenario of livestream shopping, consumers' competitive arousal negatively affects the relationship between anchors' professionalism and presence.
Our research model is represented in Figure 1.

Data collection and sampling
The sample of the survey was limited to customers who had livestream shopping experiences to investigate the influence of anchors' professionalism on consumers' impulse buying intention. First of all, two pilot tests were conducted on 3 April 2020 and 27 April 2020, targeted at 50 participants and 150 participants, respectively. Then adjustments were made to clarify survey questions. In terms of the collection of empirical data, the online survey on the wenjuanxing website (www.wjx.cn) lasted approximately two weeks from 15 May 2020 to 30 May 2020 with the hyperlink published on social media such as WeChat to invite the participants to complete the questionnaire online. There were 374 of 483 responses were collected. The valid rate of questionnaire collection was 77.43%, with a gender distribution of 19.79% and 80.21% for men and women, respectively. The main range of the age group was mainly between 18 and 35 years old, accounting for 92.51% of the total sample. The distribution between 36 and 50 years old accounting for 4.28% and age more than 50s accounting for 1.07%. The distribution of the highest degree was mainly concentrated in the bachelor's degree and master's degree, which account for 49.20% and 32.35% of the total sample, respectively. The distribution in senior high school and below accounting for 3.48%. And also the samples of PhD degree accounting for 7.49%. As for the distribution of positions, it was mainly concentrated in students and enterprise employees. And also, staff of colleges and institution, freelancers, government employees, housewife/househusband and retirees accounted for 5.88%, 6.68%, 5.08%, 1.60% and 0.53% respectively. Additionally, the distribution of the average monthly disposable income was less than 5000 yuan, accounting for 83.95% of the total sample. According to the QuestMobile 2021 'her economy' insight report (QuestMobile 2021), in different live broadcast platforms, female users account for more than 80% and female users under the age of 35 are the main live broadcast e-commerce user group. Therefore, considering the fact most of the consumers who love livestream shopping are young women, the distribution of samples collected in this study is in line with the reality. The detailed information about the sample is shown in Table 1.

Measurement development
The optimisation of the survey items was based on the previous literature. The questionnaire is composed of 20 questions as shown in Table 2, slightly modified from the previous literature to make it more applicable to the research background of livestreaming e-commerce. For example, the items for measuring anchors' professionalism were adapted from those in the study by Ohanian (1991) (OHANIAN 1991). The items for measuring consumers' sense of presence were adapted from those in the studies by Barfield et al. (1995), Hassanein and Head (2007), and Wang et al. (2009) (Barfield, Zeltzer, andSheridan et al. 1995)(Hassanein and (Wang, Zheng, and Zhou 2009). The items for measuring consumers' competitive arousal were adapted from those in the study by Nichols and Flint (2010) . And the items for measuring consumers' impulse buying intention were adapted from those in the study by Beatty (1998) and Vijayasarathy (2002) (Beatty 1998) (Vijayasarathy 2002). A five-point Likert scale was applied to each item, with one being 'strongly disagree' and five being 'strongly agree'.  Table 3 shows, the means, standard deviations (SD), and correlations of the study variables. In this study, the stability and reliability of the results were evaluated based on Cronbach's alpha coefficient. As shown in Table 3, the overall Cronbach's alpha coefficient of the questionnaire was 0.893 (higher than 0.8), indicating that the questionnaire's overall stability is good. At the same time, Cronbach's alpha coefficients of all dimensions in the questionnaire were over 0.7, which belongs to the acceptable level. Then the exploratory factor analysis was conducted using SPSS 22.0, showing that Kaiser-Meyer-Olkin (KMO) was 0.911, more than 0.8, and the questionnaire's factor analysis is suitable. Four principal components were finally extracted with a cumulative interpretation degree of 72.995%, which is consistent with the original research model, indicating that the questionnaire design is reasonable.

Reliability and validity tests
According to Table 4, the average variance extracted (AVE) and combined reliability (CR) of variables were greater than 0.5 and 0.6, respectively, indicating that the good aggregation validity of the questionnaire.

Comparison of measurement models
In order to assess the discriminant and convergent validity of the constructs, a number of confirmatory factor analyses (CFAs) were performed using the maximum-likelihood estimation method (Kline 1998). The expected four-factor solution (anchors' professionalism,  (Kraimer et al. 2012). In the light of the high correlation between consumers' impulse buying intention and consumers' competitive arousal, as well as the high correlation between consumers' impulse buying intention and presence, more parsimonious models were tested to compare them with the four-factor solution. In Model 2, a three-factor model, consumers' impulse buying intention was combined with consumers' competitive arousal. In Model 3, a three-factor model, consumers' impulse buying intention was combined with presence. And Model 4 is a onefactor model. The results of the Chi-square tests demonstrate that the fit of the four-factor model is significantly better than that of other models. Therefore, the four-factor can provide basic evidence for discriminating the validity of the results. Furthermore, the factor loads of all projects in the four-factor model were found to be significantly In the scenario of livestream shopping, I think the products recommended by anchors are reliable. Ohanian (1991) In the scenario of livestream shopping, I trust the anchors.
In the scenario of livestream shopping, I think the anchors have professional skills. In the scenario of livestream shopping, I think the anchors have a good understanding of the products they recommend. Presence In the scenario of livestream shopping, I feel as if the products are in front of me. Barfield et al. (1995) Hassanein and Head (2007) Wang, Zheng and Zhou (2009) In the scenario of livestream shopping, I feel like I am shopping in a real shopping mall. In the scenario of livestream shopping, I feel like I'm dealing with people face to face. In the scenario of livestream shopping, I feel like the anchors and I know each other's thoughts. consumer competitive arousal In the scenario of livestream shopping, I feel like I'm competing with other people to buy products. Nichols and Flint (2010) In the scenario of livestream shopping, when I successfully snap up a product, I feel like I've won over the others. In the scenario of livestream shopping, I feel anxious about whether I can buy the product.

Consumers' impulse buying intention
When an anchor recommends products, I want to own one immediately.

Beatty (1998) Vijayasarathy (2002) When an anchor recommends products, I think that's what I want to
buy. In the scenario of livestream shopping, I always find many products that I do not plan to buy before, but I want to buy after the anchors recommend them. Consumers' impulsive purchase intention 3.135 0.875 0.554** 0.587** 0.654** 0.793 a n = 374. bp < 0.05, **p < 0.01, ***p < 0.001; alpha coefficient are on the diagonal in bold. correlated with their respective underlying constructs (p < 0.001), which provides evidence for the convergence validity. The results of the measurement model comparison are shown in Table 5.
In addition, based on the four-factor model, the maximum likelihood method was used to examine the normal distribution of the measured variables, and AMOS 22.0 was used to estimate the parameter. All fitting levels do not need to gain higher values than the standard (Hair 2009). The overall results of the fitting evaluation are as follows: χ 2 = df ¼ 2:478 (≤3), RMSEA = 0.063 (0.05 ~ 0.1 or below: acceptable), CFI = 0.963 (0.9 or more), NFI = 0.939 (0.9 or more), IFI = 0.963 (0.9 or more), and TLI = 0.952 (0.9 or more). According to the structural models of this study, most values were close to or completely in accordance with the standard values. That is to say, the model fitting is sufficient.

Test of hypotheses
The regression analysis (Cohen, Cohen, and G et al. 2014) was used to test the hypotheses in this study. The results are presented in Tables 6 and 7. For Hypothesis 1, anchors' professionalism in the scenario of livestreaming e-commerce was found to be positively correlated with consumers' impulse buying intention (Model 4, β = 0.571, p < 0.001). Thus, Hypothesis 1 is supported. The result remained unchanged when controlling the variables, including gender, age, highest degree, position, and income.
For Hypothesis 2, it was found that there was a positive correlation between presence and consumers' impulse buying intention (Model 5, β = 0.582, p < 0.001), and thus Hypothesis 2 is supported. For Hypothesis 3, a positive correlation between anchors' professionalism and presence was found (Model 2, β = 0.544, p < 0.001), and thus Hypothesis 3 is supported. For Hypothesis 4, the mediation of the impact of anchors' professionalism on consumers' impulse buying intention was found to be realised by consumers' sense of presence. To be more specific, the bootstrapping programme was used to construct a 95% confidence interval for the bias correction of the indirect effect, with 1000 random samples taken from the complete sample for the replacement. The confidence interval for presence (0.176; 0.357) excluded 0. Thus, the hypothesised mediation is supported (Shrout and Bolger 2002).
For Hypothesis 5, it was found that consumers' competitive arousal moderated the interaction between anchors' professionalism and presence. As shown in Table 6, the coefficient of the interaction between anchors' professionalism and consumers' competition arousal was significant (Model 3, β = −0.092, p < 0.05), and thus hypothesis 5 is supported. Figure 2 displays the simple slopes for this moderation, demonstrating that the consumers whose sense of competition was not evoked (M-1SD), anchors' professionalism had a significant effect on their perceived presence (simple slope = 0.582, t = 0.791, p < 0.001), and for the consumers whose sense of competition was evoked (M + 1SD), anchors' professionalism also had a significant effect on their perceived presence (simple slope = 0.438, t = 5.571, p < 0.001).
Moreover, to test the mediating role in the hypothesis, hypothesis 4 was tested following the same procedure. It was found that the confidence interval for the mediation excluded 0 at the low (0.149; 0.298), medium (0.123; 0.281), and high (0.070; 0.287) levels of the moderator, respectively. Therefore, the hypothesised mediation is supported (Shrout and Bolger 2002). Notes, Comparisons are made with model 1. Model 1: Four-factor model. Model 2: Three-factor model with consumers' impulse buying intention and consumer competitive arousal loading on a single latent factor. Model 3: Three-factor model with consumers' impulse buying intention and presence loading on a single latent factor. Model 4: One-factor model. RMSEA = root mean square error of approximation; CFI = comparative fit index; SRMR = standardised root mean square residual.

Robustness test
Considering that the original empirical research sample group was dominated by females, therefore, in order to test the robustness of the results, we distributed questionnaires for research again between 5 January 2022 and 5 February 2022, and only the male group was considered at the robustness testing stage. 339 valid questionnaires were returned. If the results of the study are still significant, it further indicates the robustness of the findings of this study. Regarding the basic distribution of this part of the sample, the main range of the age group was mainly between 26 and 35 years old, accounting for 69.62% of the total sample. The distribution of the highest degree was mainly concentrated in the bachelor's degree, which account for 74.63% of the total sample. As for the distribution of positions, it was mainly concentrated in enterprise employees, which account for 75.52% of the total sample. Additionally, the distribution of the average monthly disposable income was mainly concentrated between 5001 yuan and 10,000 yuan, accounting for 48.97% of the  total sample. Through the regression analysis of the results of this research, the results of the study are shown in Table 8. In the live e-commerce scenario, the professionalism of the anchor is still positively correlated with the impulse purchase intention of consumers. The mediation effect between anchor's professionalism and consumer's impulse purchase intention still holds. Meanwhile, the negative moderating effect of consumers' emotional arousal is significant. In conclusion, the main findings of this study remain significant through another research and analysis of the male sample. There is no significant change in the direction and dimensionality of the coefficients of the main variables. Therefore, the conclusions presented in this study can be considered robust.

Discussion
In this study, a literature review was combined with empirical research to explore the impact of anchors' professionalism, presence, and consumers' competitive arousal on consumers' impulse buying intention in the scenario of livestreaming e-commerce. In the literature review identified the direct influence of the professionalism of information providers on the persuasiveness of information (Kelman 1961) has been identified, which is the key factor that affects the consumer's purchase behaviour according to Chen and Zhang (2019) (Chen and Zhang 2019). The theoretical model is based on the information source theory in the research field of consumer behaviour. Moreover, the research gap in the field of consumer behaviour in livestreaming e-commerce has been narrowed. Most previous researches have indicated that product characteristics are the main factors affecting consumers' purchase intention, but the results of this study have proven the importance of anchors in livestreaming e-commerce. In terms of empirical research, the theoretical research models for consumers who had livestream shopping experiences before have been established and then verified. The results have shown that anchors' professionalism is a significant factor that affects consumers' purchase intention in livestream shopping. In addition, consumers' competitive arousal has been found to have an impact on the relationship between anchors' professionalism and presence. As a result, for practitioners in livestreaming e-commerce, it is important to evoke a sense of competition among consumers.
The concept of anchors' professionalism similar to the salesperson's professionalism in the mall implies the importance of the salesperson's familiarity with the products being sold and sales skills. Particularly, it is important to evoke consumers' impulse buying intention in livestreaming e-commerce. In this study, the pertinent factors that contribute to the success of anchors' recommendations, including anchors' professionalism, presence, and consumers' competitive arousal have been explored. The research results have shown that both anchors' professionalism and presence are positively correlated with consumers' impulse buying intention. At the same time, anchors' professionalism positively affects the perceived presence of consumers. And presence plays a mediating role in the interaction between anchors' professionalism and consumers' impulse buying intention. Moreover, consumers' competitive arousal is negatively correlated with the interaction between anchors' professionalism and presence. As one of the common ways of promotion, consumers' competitive arousal has the advantage of evoking consumers' impulse buying intention. When consumers focus on snapping up products during livestream shopping, they are concentrated on whether they can successfully buy the product rather than whether the anchor's recommendations are professional. Therefore, for the practitioners in livestreaming e-commerce, especially anchors, it is crucial to develop professionalism to successfully motivate consumers to buy something. However, appropriate measures should be considered to stimulate consumption because of the differences in people's abilities. For example, some strategies or games can be designed to stimulate consumers' sense of competition during livestream shopping, thus enhancing their shopping intention and improving economic benefits.

Implications for theory
The current study has several theoretical and practical contributions. First of all, it introduced a new research scenario, that is livestream shopping, and explored the relationship between anchors, commodities and consumers in the scenario of livestream shopping. The interactive relationship among anchors' professionalism, presence, consumer competitive arousal, and consumers' impulse buying intention have been comprehensively revealed in this study. The existing studies rarely focus on the arousal of consumers' sense of presence and impulse buying intention based on anchors' characteristics. But as the key information disseminator, anchors will find it not enough to stimulate consumers' loyalty in the scenario of livestreaming e-commerce through personality traits such as humour. Only when consumers feel that anchors have professional competences and understand their needs well can it be possible for anchors to arouse consumers' sense of identity with anchors and then enhance their trust in the products recommended, as well as impulse buying intention. As an emerging shopping mode with diverse real-time interactions between anchors and consumers, livestream shopping subverts the way of consumers' selection of commodities by themselves in traditional online shopping. The emergence of anchors has increased new development opportunities for e-commerce. Since videos can spread information more efficiently than texts, sounds, or pictures, livestream shoppers can obtain more information than traditional online shoppers. The efficient interaction between anchors and consumers in a relatively limited space is more in line with the information acquisition habit of modern consumers. On the one hand, consumers in livestream shopping do not need to sift tons of products. Instead, anchors replace consumers to complete the selection process of products, and consumers only need to choose the products recommended by anchors, which is a more efficient process of shopping. On the other hand, livestream shopping is a form of entertainment. For consumers, shopping is also a way to kill time, during which they can achieve psychological satisfaction by participating in livestream shopping. Consumers can interact with both anchors and other consumers. The dual entertainment experience of social interaction and shopping can be realised without leaving home, stimulating consumers' sense of immersion in the livestream and arousing their impulse buying intention. Second, the moderating effect of consumers' competitive arousal on the interaction between anchors' professionalism and presence has been explored in this paper. Previous studies have paid little attention to the arousal of consumers' sense of competition, especially in the specific scenario of livestream shopping, and the role of consumer's competitive arousal in consumer behaviour has not yet been clarified. This study has introduced consumer competitive arousal into the livestreaming e-commerce scenario and found an interesting phenomenon that consumers competitive arousal is negatively correlated with the interaction between anchors' professionalism and presence. This finding not only enriches the literature on consumer competitive arousal but also provides a new perspective for research on consumer behaviour.

Implications for practice
This study contributes to the following practical aspects. First, technological convenience is only one of the antecedent variables that stimulate presence. As the subject who has direct communication with consumers, the anchor has a greater impact on consumer emotional arousal. With the continuous development of the livestreaming e-commerce market, consumers will have increasingly high requirements for anchors. The ability to explain commodities clearly is a professional ability that anchors must possess. In order to ensure the sales volume, re-purchase rate, and consumers' loyalty, anchors must constantly develop their professionalism, and the threshold of the profession needs to be gradually standardised. Second, arousing a sense of competition is one of the most common ways to stimulate consumption in marketing, which works equally well in the livestream shopping scenario. Developing the anchor's professionalism is based on high costs of time and training, whereas the costs of setting up the competition mechanism and creating a competitive atmosphere in the process of livestreaming marketing are relatively low. To some extent, the mechanism can also distract consumers from their attention to anchors' professionalism and is beneficial to better marketing effects and higher economic profits. Therefore, creating a competitive atmosphere in the process of livestream shopping is an effective way to stimulate consumption. To sum up, for anchors, competitive elements can be properly used to stimulate consumers' impulse buying intention, but the cultivation and development of professionalism cannot be ignored.

Limitations and future research
Apart from theoretical and practical implications, this study has some limitations related to the research sample and content. Firstly, affected by the availability of the sample, the age group for the sample is concentrated in the range from 28 to 35 years old, which is a relatively young group and may lead to some biases in the results of the study. Thus, the sample coverage should be expanded in future studies. Secondly, anchors' professionalism was taken as an independent variable affecting consumers' impulse buying intention, but there are other dimensions related to anchors such as personal attractiveness, which may also have an impact on consumers' impulse buying intention. In the future, the influence of other characteristics of anchors on consumer behaviour can be explored from a more comprehensive dimension. Finally, in the future, we hope researchers can consider more control variables and adopt different regression models to enhance the robustness of our study.