In the history of the Olympic Games, in addition to the site selection, preparation and hosting of previous events, the host country will also carefully design and select the Olympic mascot image to highlight national characteristics, enhance the promotion of the events and drive the Olympic economy [1]. As a medium symbol for many large sports events, the official mascot is increasingly important and has become an indispensable part of large-scale event management, especially revenue and legacy management [2]. In the academic world, the Olympic mascots have been in the spotlight as an important part of the event's legacy. Existing research discussed the Olympic mascots from the perspectives of mascot image design, national image display, economic value drive and cultural communication [3]. As early as 2014, by studying the spread and influence of Olympic mascots from 2006 to 2012 in the electronic age, Knight et al. believed that mascots would play a more extensive and far-reaching role in the future event management, and emphasized the need to pay attention to the correlation between mascots and the interaction of the spectators. Palladino et al. [4] adopted the online experimental method to study the influence mechanism of Olympic mascot design elements on fans' perception, attitude and purchasing intention, and found that anthropomorphic design is more likely to attract the spectators and stimulate their purchasing behavior. It is worth emphasizing that the non-present spectators' motivation to participate in special events is more significant than that of the on-site paying spectators, but the non-present spectators’ participation mode and sense of participation are not paid much attention in the event marketing. Overall, it remains unknown how personification affects the social interactions between people, especially non-present spectators. Existing relevant researches adhere to a multidisciplinary perspective and mainly adopt qualitative methods, moreover, there are many discussions on the impact of the economic value of Olympic mascots, and insufficient attention is paid to the social value of sports mascots. In addition, based on quantitative analysis, some literature explored the influence of personification of mascot on purchasing behavior, but there are few studies on the influence mechanism of mascot on social interaction such as attracting spectators’ love, communication and transmission. The unexpected popularity of the 2022 Beijing Winter Olympics mascot has become a highlight of the Winter Olympics. With the endorsement of celebrities from all over the world, including the Prince of Monaco, Bing Dwen Dwen's related franchised goods were quickly sold out. Countless netizens have shared and spread photos on the Internet in the form of text, tags or GIFs, and Bing Dwen Dwen's Weibo topic has been read 6.03 billion times [5]. In this process, there are not only the spectators at the scene of the competition, but also the people from all over the world who did not "visit" the scene. This is the most popular mascot in the history of the Olympic Games, and such a sensational social driving effect is worth in-depth discussion and analysis. In view of this, this paper takes the anthropomorphism of the Olympic mascots as the starting point, with Bing Dwen Dwenas an example, and uses the social information processing theory to explore the influence mechanism behind the Olympic mascots' promotion of the social interaction of non-present spectators, in order to provide a new perspective for the research on the mascots of sports events. It also provides theoretical guidance for the design, commercial development and communication channels of future sports event mascots.

Anthropomorphism of Olympic Mascots

The word Mascot comes from the French word Mascotte and is defined by Merriam-Webster dictionary as "a person, animal, or object adopted by a group as a symbolic figure especially to bring them good luck". A mascot is usually designed to symbolize a specific organization, event, or group [6]. The first officially recognized Olympic mascot was Vardy, a dachshund who represented the Munich Summer Games in 1972. Since then, every Olympic Games has had its own mascot, which has been used effectively as a marketing tool and has been derived from a range of consumer products such as dolls, handpieces, stationery, etc. Knight et al. [7] studied the influence of Olympic mascots from 2006 to 2012 from the perspective of semiotic analysis, suggesting that the interaction between mascots and the spectators should be paid attention to in the future. Anthropomorphism refers to the idea that people ascribe specific human characteristics, personality, behavior, thinking and mental state to real or imagined non-human entities, including animals, natural forces, religious roles, technological tools, etc., so as to realize the interaction between the two [8]. This is a common way of thinking, such as people communicating with animals and plants, empathizing with beloved dolls, seeing smiling faces in the clouds, etc. Some scholars have conducted researches on anthropomorphic motivation and proposed the classic theory of three factors of anthropomorphism, that is, motivating subject knowledge, efficacy motivation and social motivation can all affect anthropomorphic tendency. The research on anthropomorphic marketing is more extensive, mainly divided into product anthropomorphic and brand anthropomorphic two directions. For example, anthropomorphism can virtually bring products or services closer to customers. When consumers personify a brand, it helps to produce behaviors that contrast or assimilate with the brand image [9]. With the continuous improvement of scientific and technological means, artificial intelligence and anthropomorphism are further integrated, and products such as intelligent service systems and intelligent service robots promote the innovative development of the service industry [10]. According to previous research, it can be seen that the application of anthropomorphic means is increasingly mature and effective. As an important image representative of major events, characteristic cities and enterprises, Mascot is also deeply affected by anthropomorphic factors. For example, Tateishi et al. [11] found that mascots with simple design and anthropomorphic features were more interesting to the spectators. [12] Experimentally demonstrated that mascots with anthropomorphic design have a positive impact on consumers' brand purchase intention. These research results comprehensively express the important role of anthropomorphic design of mascots, and also point out that it is worth exploring to study the Olympic mascots from the perspective of anthropomorphic design. Wang and Xie [13] summarized the research on anthropomorphic marketing and concluded that anthropomorphic marketing is an intermediary to establish contact with non-human entities to achieve individual social needs, and consumers mainly accept anthropomorphic information sources such as language, image, appearance and body movements. It can be seen that the anthropomorphic design of Olympic mascots mainly conveys anthropomorphic information to consumers through image appearance and body movements, and is displayed to the public through various forms such as humanoid doll play, film promotion, live competition and exclusive expression, so as to promote the connection between events and participants and enhance the event experience of participants.

Social Interaction

Social interaction refers to the dynamic process of interaction and interdependence between social individuals or groups based on information dissemination to meet various needs [14]. In the process of social interaction, individuals are always the subject of interaction, and the main purpose of interaction is to realize interpersonal information communication, the interaction of the two sides' actions and the conversion of specific resources. From the perspective of the core elements of social interaction, events are the most abundant form of social interaction. Special events are generally considered to be an important platform that can bring people together for social interaction and influence different types of social interaction among participants [15]. Social interaction at events takes different forms, such as family reunions, participation in local festivals, learning skills, and meeting new people. Large-scale sports events, such as the Olympic Games, World Cup, etc., contain rich information elements and various forms of interaction, and attract a large number of spectators. During the holding period, professional performance, beautiful peripheral products, unexpected results, etc., will become hot topics and become a driving force for spectators to generate social interaction.

The Effect of Anthropomorphism on Social Interaction

Based on the combing of relevant literature above, Information Processing Theory is used in this research. Social Information Processing Theory means that participants will self-decode and process signals sent by objects, and then obtain information to influence their own psychology and behavior. Under different circumstances, participants will actively identify important clues useful to themselves to meet the needs of social interaction [16]. Reports from various official news outlets and social media, especially short video platforms, showed that the 2022 Beijing Winter Olympic Games attracted much attention from non-present spectators. In addition, they actively share and create Bing Dwen Dwen emojis on social media platforms and other channels. Therefore, it is speculated that anthropomorphic mascots are more likely to win the favor of non-present spectators, and people will have a sense of closeness and sharing desire due to their human-like features, thus encouraging them to take the initiative to learn more about the mascot and related events. Based on this, this paper puts forward the following hypothesis: H1: The anthropomorphic level of mascots of large-scale sports events has a positive impact on the social interaction of non-present spectators. Compared with the Olympic mascots with a low anthropomorphic level, the Olympic mascots with a high anthropomorphic level can enhance the social interaction more significantly, and have a more significant impact on the social interaction among unfamiliar groups.

The Mediating Role of Perceived Loveliness

Perceived cuteness refers to the expression of positive emotions that can induce individual affection or compassion. In the current research, the ways of designing cuteness include impression and interaction. In terms of impression, designers integrate "cuteness" with products through color, movement, size, etc. Studies showed that baby-like or feminized appearance designs are more likely to inspire spectators’ affection [17]. In the interactive aspect, cuteness can be reflected in attitude, language expression and personality. For example, et al. [18] analyzed the loveliness of mascots through the dimensions of emotion, language and design, among which, the measurement of emotion dimension mainly came from the anthropomorphic design of mascots. In addition, Helen and Saaksjarvi [19] believed that as the degree of perceived cuteness increases, consumers will show stronger interest or love, thus displaying multi-dimensional interactive behaviors. Cute products can also effectively promote indulgent consumption and reduce consumers' regret after purchase, and enable them to have a higher sense of engagement and experience [20]. During the course of the games, when the spectators have affection or sympathy for the Olympic mascots, the behavior of contacting and interacting with them will increase; and when people with similar feelings gather together, it also provides an opportunity for individuals to interact with unfamiliar groups. Based on this, this paper proposes the following hypothesis:

H2a: Anthropomorphic level of mascots of large-scale sports events has a positive influence on perceived cuteness.

H2b: Perceived cuteness has a positive impact on the social interaction of non-present spectators, and its impact on the interaction with unfamiliar groups is more significant.

H2c: The anthropomorphic level of mascots of large sports events influences the social interaction of non-present spectators through the mediating effect of perceived cuteness.

The Mediating Effect of Psychological Distance

The concept of psychological distance was born in the field of international trade. Later, psychologists believe that it can be extended to the field of psychology to explain macro issues such as people's cognitive unity towards things [21]. Lee et al. [22] proposed to study the role of psychological distance in consumers' psychological perception from an individual microscopic perspective, and then psychological distance was gradually widely used in the research of marketing communication. Psychological distance refers to the subjective perception that people or things are close or far away from them, which is subdivided into four levels of perception: time (earlier or later), space (closer or farther), social (familiar or unfamiliar) and hypothesis (certainty or uncertainty). In recent years, many scholars are more inclined to explore the relationship between the interaction effect of these dimensions and individual behavior, and the distance of psychological distance will affect the change of individual behavior. When a non-human entity is personified, the similarity with human will be increased, thus shortening the psychological distance between human and the entity [23] In large-scale sports events, mascots, as the "spokesperson" of the event image publicity, act as a bridge between the event and the spectators, narrowing the distance between the spectators and the event in an interesting way, with the purpose of creating interactive experience and attracting more spectators’ attention. Based on this, this paper proposes the following hypothesis:

H3a: The anthropomorphism of mascots of large-scale sports events negatively affects the perceived psychological distance of the spectators, that is, the higher the anthropomorphic level of the Olympic mascots can shorten the psychological distance more effectively.

H3b: The psychological distance perceived by the non-present spectators negatively affects their own social interaction, and the shorter the psychological distance perceived by the non-present spectators, the easier it is to have social interaction with unfamiliar groups.

H3c: The anthropomorphism of mascots of large sports events affects the social interaction of non-present spectators through the mediating effect of psychological distance.

The Moderating Effect of Self-Expression

Self-expression is the "declaration of self-characteristics", the process of conveying individual thoughts, preferences and feelings to others, and a basic human need, with the purpose of shaping oneself in the eyes of others [24,25] Individuals will habitually express themselves in daily life, adopt various behaviors to meet this universal need, and elicit feedback from others through self-expression [26]. Numerous marketing studies have proved that consumers use products and brands as expression tools to symbolically construct and express their own self-concept [27-29] Similarly, in the field of psychological research, choice is the behavior of an individual's internal expression of the explicit and concrete [30-33], choosing and investing oneself in a particular course of action, actively maintaining self-concern in order to defend the outcome of self-expression, and demonstrating a fondness or preference for the chosen option [25,34]. According to Self-Determination Theory (SDT), self-expression is regarded as a kind of autonomy or intrinsic motivation. When motivated by emotion, perception and other intrinsic motivations, people choose to take the initiative to participate in special events or complete specific tasks to achieve self-expression needs [35,36] found that consumers tend to choose corresponding product designs in order to show themselves and achieve uniqueness. Similarly, Olympic mascots, as marketing products for large-scale events, show their uniqueness in design, publicity, image, exhibition and other aspects. The anthropomorphic design of the Olympic mascots can awaken the spectators' inner lovely perception, narrow the psychological distance between the spectators and the events, and also inspire the majority of the spectators to desire to show their self-expression needs such as loveliness, sincerity and closeness in communication with others through the mascot products. On the other hand, the stronger the spectators' need for self-expression, the more willing they are to take the initiative to use the object as a medium for social interaction. Based on this, this study proposes the following hypothesis:

H4a: Non-present spectators’ self-expression positively moderates the influence relationship between perceived likability and social interaction. The stronger the non-present spectators' self-expression, the stronger the influence of their perceived cuteness on social interaction.

H4b: The non-present spectators' self-expression positively moderates the relationship between psychological distance and social interaction. The stronger the non-present spectators' self-expression, the stronger the influence of their perceived psychological distance on social interaction.

To sum up, the hypothetical model of this study is shown in Fig.1.

Figure.1: Research hypothesis of this research.

Experiment 1: Main Effect of Mascot Personification on Social Interaction with Non-Present Spectators

In experiment 1, a single-factor at two-level (high anthropomorphism vs. low anthropomorphism) intergroup experiment was designed to test the effect of anthropomorphism level of mascots of large-scale sports events on social interaction of non-present spectator that is to test H1. In order to ensure the effectiveness of experimental materials in controlling independent variables, a pre-experimental study was conducted before the formal experiment began.

Pre-experiment

In this stage, the online platform credemo was used to design and deliver the scenario experiment of "text + pictures". The participants were asked to watch pictures and videos and complete the corresponding feedback questions, which aims to prove the effectiveness of this experiment on the anthropomorphic level manipulation of the mascot. By asking "Have you ever been to the 2022 Beijing Winter Olympic Games?"50 non-present spectator members were selected and recruited as participants, and randomly assigned to two groups, and different experimental materials were shown to the two groups respectively. After the presentation, the subjects were asked to evaluate the anthropomorphic design level of the materials. The independent variables were measured using the anthropomorphic design scale of Kim and McGill [37], with a total of 5 items, such as "I think this mascot looks like a human", "I think this mascot has human-like characteristics", and "I think this mascot has its own consciousness". Based on the dependent variable, Marques et al. [38] combined the practice of co-creation, group socialization and interactive ritual chain, and developed the Event Social Interaction Scale (ESIS). This scale can scientifically measure the impact of holding an event on participants' social interaction, specifically considering the two levels of participants' interaction, perception and behavior with known groups and unfamiliar groups, consisting of 14 items. In addition, gender, age, education and family status were used as control variables. In order to identify and eliminate invalid data, "difficulty of situation imagination, degree of semantic ambiguity, degree of picture clarity" and other screening items were added according to Bearden et al. [39], Dabholkar and Bagozzi [40]. Likert 5-point scale method (1 represents "strongly disagree", 5 represents "strongly agree") was adopted for all the above measurements. Through the independent sample T-test results, it was found that the mean value of the high anthropomorphic experimental group was significantly higher than that of the low anthropomorphic experimental group (MHAL=4.392, MLAL=2.808, t=-13.264, p<0.001). The independent variable (anthropomorphic level) was successfully and effectively manipulated by the experimental materials, and a formal experimental study could be conducted on this basis.

Experimental Manipulation and Procedure

Based on the pre-experiment process and operation results, the formal experiment was conducted using credemo. The procedure of experiment 1 was consistent with the pre-experiment. 250 non-present spectators were recruited as experiment participants through the online platform question screening, and they were randomly assigned to the two experimental groups and given material stimulation with different levels of anthropomorphism. After the experiment, they were asked to fill in the anthropomorphic design level measurement scale and the ESIS, both of which are Likert 5-point measurement. Through data screening, 37 invalid questionnaires were eliminated due to wrong filling, missing filling and failing to pass the identification items. The effective sample size of the experiment was 213, with an effective rate of 85.2%, 103 males and 110 females, among which 48.8% were from 18 to 28 years old, 36.6% were from 28 to 38 years old and 15% were from other ages. Educational background was mainly distributed in college, undergraduate (77.9%) and postgraduate (17.4%), and occupation was mainly distributed in students (22.5%), enterprise managers (including grassroots and middle and senior managers) (33.8%), and professionals (such as journalists/teachers/doctors/lawyers) (19.7%). The level of anthropomorphism was significantly different between the two experimental groups (MHAL=4.241, MLAL=2.585, t=-31.452, p<0.001), indicating that the variable manipulation was successful.

Result Analysis and Discussion

According to the results of one-way variance test, there were significant differences in perceived cuteness between the two experimental groups (MHAL=4.184, MLAL=3.121, F (1,244) =233.4,p<0.001), that is, anthropomorphism initiated the non-present spectators' perceived evaluation of the mascot's cuteness, and H2a was verified. Consistent with the results of experiment 1, social interaction was significantly different between the two experimental groups (MHAL=4.145, MLAL=3.298, F (1,244) =412.6,p<0.001).

In order to further clarify the relationship between variables, experiment 2 established a general linear regression equation with social interaction as the dependent variable and perceived cuteness as the independent variable, and took demographic variables such as gender, age, education and occupation as the control variables. The results showed that perceived cuteness (β=0.827, p<0.001) had a significant influence on social interaction coefficient. In addition, in the influence of perceived cuteness on social interaction, the interaction degree between non-present spectators and unknown group (β=0.562, p<0.001) is higher than that between non-present spectators and known group (β=0.265, p<0.01), so H2b is verified. Finally, the study further tested the mediating effect of perceived cuteness. Specifically, the low anthropomorphic level was coded as 1 and the high anthropomorphic level was coded as 2 and the SPSS PROCESS program was adopted, model 4 (mediating model test) was selected, and the number of Bootstrapping was set to 5000. The results showed that anthropomorphic level had a significant direct effect on the social interaction of non-present spectators (βdirect effect=0.845, LLCI=0.667, ULCI=1.024, interval excluding 0) (see Tab.1.). The confidence interval of perceived cuteness is 0 free and reaches the significant level(β indirect effect =0.248, LLCI=0.129, ULCI=0.363), indicating that anthropomorphic level can influence the social interaction of non-present spectators through the mediating effect of perceived cuteness, and H2c is supported.

Table 1: Test results of Bootstrap mediation model (mediating effect of perceived cuteness).

Experiment 3: Mediating Effect of Psychological Distance

In experiment 3, the materials and procedures are basically the same as in experiment 1, and the measurement of psychological distance is added in order to test its mediating effect in the main effect, that is to test H3. Using the psychological distance measurement scale of Ren et al. (2014), there are three items such as "this image makes my heart closer to the event" and "this image reduces my stranger feeling to the event". The higher the measurement score, the more effective the image is in narrowing the distance between the participants and the event, and the shorter the psychological distance.

(1)Experiment manipulation and procedure

In experiment 3, recruitment requirements were set on the credamo platform, excluding participants who had participated in the pre-test, experiment 1 and experiment 2, and a total of 250 non-present spectators were recruited. After deleting 16 invalid data, the effective sample size was 234, including 120 males and 114 females and the effective rate was 93.6%. Among them, 64.1% were from 18 to 28 years old, 26.92% were from 28 to 38 years old, and 8.98% were from other ages. Education is mainly distributed in college, undergraduate (54.7%) and postgraduate (38.0%), while occupation is mainly distributed in students (29.5%), enterprise managers (including grassroots and middle and senior managers) (26.5%), and professionals (20.9%), such as journalists/teachers/doctors/lawyers. The level of anthropomorphism was significantly different between the two experimental groups (MHAL=4.054, MLAL=3.18, t=-13.020, p<0.001), indicating that the variable manipulation was successful. Anthropomorphism (α=0.755), social interaction (α=0.870) and psychological distance (α=0.685) all passed the internal consistency test.

(2)Result analysis and discussion

According to the results of one-way variance test, there were significant differences in psychological distance variables between the two experimental groups (MHAL=4.187, MLAL=3.036, F (1,232) =290.0,p<0.001), that is, anthropomorphism initiated the perception of psychological distance between non-present spectators and mascot, and H2a was verified. Consistent with the results of experiment 1 and experiment 2, there were significant differences in social interaction between the two experimental groups (MHAL=4.114, MLAL=3.192, F (1,232) =396.8,p<0.001). In order to further clarify the relationship between variables, experiment 3 established a general linear regression equation with social interaction as the dependent variable and psychological distance as the independent variable, and took demographic variables such as gender, age, education and occupation as the control variables. The results found that psychological distance (β=1.487, p<0.001) had significant impact on social interaction coefficient, and, among the influences of this variable on social interaction, the interaction degree between non-present spectators and unknown group (β=1.127, p<0.001) is higher than that between non-present spectators and known group (β=0.360, p<0.001), so H3b is verified. Finally, the study further tested the mediating effect of psychological distance. Encoding the low anthropomorphic level as 1 and the high anthropomorphic level as 2, the SPSS PROCESS program was adopted, model 4 (mediating model test) was selected, and the number of Bootstrapping was set to 5000. The results showed that the direct effect of anthropomorphic level on social interaction was significant (βdirect effect=0.957, LLCI=0.695, ULCI=1.219, interval excluding 0) and the confidence interval of psychological distance is also 0 free and reaches the significant level (βindirect effect=0.546, LLCI=0.396, ULCI=0.711) (see Tab.2.), which indicated that anthropomorphism can exert an influence on social interaction through the mediating effect of psychological distance, so H3c is supported.

Table 2: Test results of Bootstrap mediation model (mediating effect of psychological distance.

Experiment 4: The moderating effect of self-expression

In this study, a hybrid design of 2 (high anthropomorphism vs. low anthropomorphism) plus 2 (strong self-expression vs. weak self-expression) intergroup and intra-group was used to examine the moderating effect of the need for self-expression, that is to test H4. The experimental materials were consistent with the previous experiments in the process. The scale designed by De Vries [35] was referred to, and the measurement of self-expression variables was added. The measurement items included six items, such as "It can let others know who I am", "It can represent what kind of person I am", and "It fits me well".

(1)Experiment manipulation and procedure

In experiment 4, screening conditions were set on the credamo platform to exclude participants who had participated in the preliminary experiment, main effect test and intermediate effect test. A total of 250 participants participated in this experiment. After eliminating 22 invalid data, the effective sample data was 228, including 102 males, accounting for 44.7%, and 126 females, accounting for 55.3%, and the effective rate was 91.2%. Among them, 61% were from 18 to 28 years old, 33.3% were from 28 to 38 years old, and 5.7% were from other ages. The educational background is mainly distributed in college, undergraduate (71.1%) and postgraduate (57%), while the occupation distribution is mainly students (61.4%), government/agency cadres/civil servants (14.5%), and professionals (such as journalists/teachers/doctors/lawyers) (10.1%). The level of anthropomorphism was significantly different between the two experimental groups (MHAL=3.877, MLAL=2.421, t=-26.098, p<0.001), indicating that the variable manipulation was successful. Anthropomorphism (α=0.871), social interaction (α=0.944), perceived likability (α=0.819), psychological distance (α=0.869), and self-expression (α=0.882) all had good reliability.

(2)Result analysis and discussion

According to the results of one-way ANOVA, there are significant differences in anthropomorphism level in social interaction (MHAL=3.746, MLAL=2.749, F (1,226) =191.991,p<0.001), and the results are consistent with the conclusions of the previous experimental research. In order to test hypothesis H4, hierarchical regression is first carried out in this study. To examine the moderating effect of self-expression on perceived cuteness→social interaction, psychological distance→social interaction, and the moderating effect of self-expression on the two mediating paths. In specific operations, demographic variables such as gender, age, education background and occupation are controlled, and independent variable (anthropomorphism), two intermediary variables (perceived cuteness and psychological distance), moderating variable (self-expression) and interaction terms between them (perceived cuteness x self-expression, psychological distance x self-expression) are put into the model. Specific results are shown in Table 3. According to the available data, the two interaction terms are significant, indicating that the moderating variable of self-expression plays a role in the path.

Table 3: Results of hierarchical regression analysis.

* p<0.05 ** p<0.01 *** p<0.001 the t value in parentheses

In order to clarify the specific effects, simple effect analysis was adopted, and after averaging the moderating variables, the difference of influence effects at different levels was verified by increasing or decreasing one standard deviation. As shown in the Fig.3. When self-expression was high, perceived cuteness had a stronger effect on social interaction (β=0.597, t=8.113, p<0.001), while when self-expression was low, the influence path was not significant (β= 0.101?t=1.385?p>0.05). Similarly, Fig.4. Shows that only when self-expression is high, the relationship between psychological distance and social interaction reaches a significant level, which is the enhancement effect (β=0.339, t=4.997, p<0.001). When self-expression is low, the relationship between psychological distance and social interaction path is not significant (β=0.111, t=1.573, p>0.05).

Figure3: Analysis of the moderating role of self-expression in the mediating effect pathway of perceived cuteness.

Figure 4: Analysis of the moderating role of self-expression in the mediating effect pathway of psychological distance.

Finally, according to the test method proposed by Edwards and Lambert (2007), Bootstrap method was used to test the mediating effect with moderating variable, and PROCESS model 14 of the expansion program of SPSS was selected, and the number of Bootstrapping was set to 5000. After sorting out the output data, Table 4 is obtained. The results showed that the direct effect of anthropomorphic level on social interaction was significant (βdirect effect= 0.174, t=3.285, LLCI =0.070, ULCI =0.278, interval excluding 0). When self-expression was weak, perceived cuteness (βdirect effect= 0.048, LLCI =-0.025, ULCI =0.123, interval containing 0) and psychological distance (βindirect effect= 0.054, LLCI =-0.019, the mediating effect of ULCI =0.135, interval 0) did not reach the significant level. In the case of strong self-expression, perceived cuteness (βindirect effect= 0.283, LLCI =0.202, ULCI =0.374, interval excluding 0) and psychological distance (βindirect effect= 0.165, LLCI =0.097, ULCI =0.239, interval excluding 0) had a significant mediating effect on social interaction. In summary, hypothesis H4a and H4b were both supported.

Table 4: Test results of mediating effect with moderating variables.

Research Conclusion

This study adopts 4 independent scenario experiments to test the hypothesis, and each experiment recruited different subjects to maintain a high degree of independence. The following conclusions are drawn: First, the anthropomorphic level of mascots of large-scale sports events positively affects the social interaction of non-present spectators, and the higher the anthropomorphic level will positively promote the interaction between non-present spectators and unfamiliar groups (except familiar groups such as family and friends). Second, the anthropomorphism of mascots can affect the non-present spectators' social interaction through the dual mediating effects of perceived cuteness and psychological distance. Third, self-expression positively moderates the effects of perceived cuteness and psychological distance on social interaction, and the stronger the self-expression, the stronger the effect of perceived cuteness and psychological distance on social interaction. Fourth, by examining the specific mediating effects, it can be seen that psychological distance plays a stronger mediating role between anthropomorphism of mascot and social interaction than perceived cuteness.

Theoretical Implications

First of all, the existing relevant studies mainly analyze the economic value and representational significance of mascots, and less explore the communication and interaction of the spectators driven by mascots. In this study, the anthropomorphic level of mascots is regarded as a factor to induce spectator's interaction, and the social significance of mascots is explored and demonstrated. It is found that the anthropomorphic design of mascots can effectively promote the spectator's interaction, especially the interaction with unfamiliar groups, and further excavates the influence of anthropomorphic design of mascots in large-scale sports events on the social interaction of non-present spectators, and expands the related research field of the impacts of events or event heritage on social interaction. Secondly, this study tested the feasibility of ESIS in the field of event research through experimental data, and demonstrated the applicability of social information processing theory in social interaction research. Based on the test of proposed hypotheses, the research explained the spectators' psychological feelings on the personification level of mascots through data. In addition, the study explored the anthropomorphic design effect of mascots through experimental method, expanding the application scope of anthropomorphic theory in the research of event heritage, and also providing literature support for experimental method in related research in the field of events.

Practical Implications

The anthropomorphism of mascot can significantly improve social interaction, especially in the aspect of socializing with strangers. Therefore, it is necessary to give full play to the anthropomorphic characteristics of mascot and improve it from the three perspectives of economy, culture and society. From the economic point of view, due to the increasingly prominent image representation effect and economic attraction of mascots, more and more sports events, cities, enterprises and companies are willing to spend time and money to create unique and exclusive mascots, and vigorously promote and publicize it during the whole process of activities. According to the research, the design of event mascot can mobilize the interaction of the spectators, which is undoubtedly positive for the organizers. Through a high degree of anthropomorphic design, the spectators can feel a stronger sense of cuteness and a shorter psychological distance, which will enhance the needs of self-expression of the spectators, and thus promote the association between the spectators and the events. Specifically, the event organizer/designer can flexibly add anthropomorphic elements (such as eyes, mouth, expressions, gestures, etc.) to the appearance design of the mascot, and professional gesture expressions and body movements can be added to the design of the language system. If there is no language system restriction, more attempts should be made to use first-person language. In order to encourage the spectators to expect to rely on the mascot to show their own characteristics, wake up the desire for self-expression, so as to more actively engage in social interaction. In this way, it also helps to promote the sales of the mascot's peripheral products by virtue of the social interaction effect. From the cultural point of view, in large-scale sports events, mascots not only account for the main proportion of the event income, but also represent the comprehensive image of the host place, host party, local culture, local residents and other aspects, and are friendly carriers to publicize the sports events, carry and disseminate local culture, and enhance exchanges between different places. Attention should be paid to the level of anthropomorphic design. To give full play to the interaction of mascots with spectators, especially non-present spectators, so as to better inherit and carry forward local culture and make the influence of local culture more lasting and far-reaching. Finally, from the social perspective, stranger socializing has always been a key problem that all social software around the world wants to solve. Whether it is on Meta, Twitter, Wechat or QQ, stranger socializing has always been regarded as an important breakthrough in the development of social software, but so far there is no good solution. Regarding stranger socializing, we can draw lessons from the research conclusion of this study. The higher the anthropomorphic level will positively promote the interaction between non-present spectators and unfamiliar groups (except family members, friends and other familiar groups). Anthropomorphic design of social software, chat content and interaction methods may improve the function of stranger socializing in social software.

In summary, this paper mainly has the following shortcomings, which need to be further studied and improved: First, this paper adopts the online random experiment platform for recruitment and experimental data collection, which may have potential deviations from offline experiments. If the research conditions permit, field experiments can be adopted in the future. Secondly, the demographic variables of the participants were taken as the control variables, and other potential factors may have an impact on the study. Finally, according to the The Uncanny Valley Theory proposed by Mori et al [40], when the anthropomorphic level of a robot rises to a certain extent, people will be afraid or intimidated, and thus their favorable impression of the robot will decline. When the mascot design is overly anthropomorphic, the positive effect of promoting social interaction may show a reverse regulation result. In future studies, the anthropomorphic level of mascots can be divided into more details and possible differential effects can be explored. At the same time, more comparative studies on mascots of large-scale sports events can be carried out to further verify the social effects of mascots.

1. Mahatmi N and Satyagraha Mascot design for the indonesian pavilion at world expo 2020//International conference of innovation in media and visual design (IMDES 2020). Atlantis Press. 2020; 20-125.
2. Chiabaut Tokyo 2020 official mascots and their contribution to the promotion of Japan’s culture and legacy. Diagoras: International Academic Journal on Olympic Studies. 2021; 5: 88-100.
3. Griggs G, Freeman I, Knight P and O Reilly N. A vision of London in the twenty-first century or just terrifying monsters: a semiotic analysis of the official mascots for the London 2012 Olympic and Paralympic Games. Leisure Studies. 2012; 31:339-354.
4. Analysis of Olympic mascot design toward attitude and purchase i International Journal of Sport Communication. 2022; 15: 190-196.
5. Gong Q. A study on the symbol representation of Winter Olympic mascots from the perspective of communication: A case study of Bing Dwen Dwen. Research on Communication Power. 2022; 6: 3-9.
6. Griggs G, Freeman I, Knight P and O'Reilly A vision of London in the twenty-first century or just terrifying monsters: a semiotic analysis of the official mascots for the London 2012 Olympic and Paralympic Games. Leisure Studies. 2021; 31: 339-354.
7. Knight P, Freeman I, Stuart S, Griggs G and O'Reilly N. Semiotic representations of Olympic mascots revisited: Virtual mascots of the games 2006-2012. International Journal of Event & Festival Management. 2014; 5: 74-92.
8. Epley N, Waytz A and Cacioppo JT. On seeing human: a three-factor theory of anthropomorphism. Psychological R 2007; 114: 864.
9. Aggarwal P and Mc Gill AL. When brands seem human, do human’s act like brands? Automatic behavioral priming effects of brand anthropomorphism. Journal of consumer research. 2012; 39: 307-323.
10. Bowen J and Morosan C. Beware hospitality industry: The robots are coming.Worldwide Hospitality and Tourism Themes. 2018; 10: 726-733.
11. Tateishi C, Inou G, Sawai H, Koyama E, KitaniY and Fujito M. The effects on the popularities and interests mascot characters' acceptability give to the general public. In 2013 Second IIAI International Conference on Advanced Applied Informatics. 2013; 373-378.
12. DEL?GOZ K and Sevtap ÜNAL. The effect of anthropomorphic mascot on the purchasing intention of consumers: An experimental study. Sosyoekonomi. 2021; 29: 229-254.
13. Wang T and Xie ZP. Review of anthropomorphic marketing r Foreign Economics and Management. 2014; 36: 38-45.
14. Godes D, Mayzlin D, ChenY, Das S, Dellarocas C, Pfeiffer B and Verlegh P. The firm's management of social interactions. Marketing letters. 2005; 16: 415-428.
15. Simons I. Events and online interaction: The construction of hybrid event communities. Leisure Studies. 2019; 38: 145-159.
16. Low GS and Mohr JJ. Factors affecting the use of information in the evaluation of marketing communications productivity. Journal of the Academy of Marketing Science. 2021; 29: 70-88.
17. Marcus A. The cult of cute: the challenge of user experience design. Interactions. 2002; 9: 29-34.
18. Novica DR, Hidayat IK and Sutrisno AA. Aesthetic of cuteness dimensions in Malang city mascot. KnE Social Sciences. 2020; 162-171.
19. Hellez K and Sääksjärvi M. Development of a scale measuring childlike anthropomorphism in products. Journal of Marketing Management. 2013; 29: 141-157.
20. Nenkov GY and Scott ML. So cute I could eat it up: Priming effects of cute products on indulgent consumption. Journal of Consumer Research. 2014; 41: 326-341.
21. TropeY and Liberman N. Construal-level theory of psychological distance. Psychological review. 2010; 117: 440.
22. Lee JK. The effects of familiarity and similarity on psychological distance toward trustworthiness of an online company. Michigan State University.2005.
23. Liviatan I, TropeY and Liberman N. Interpersonal similarity as a social distance dimension: Implications for perception of others’ actions. Journal of experimental social psychology. 2008; 44: 1256-1269.
24. Goffman E. The presentation of self in everyday life. 1959. Garden City NY 259. 2002.
25. Kim HS and Sherman DK. Express yourself": culture and the effect of self-expression on choice. J Pers Soc Psychol. 2012; 92: 1.
26. Dominick JR. Who do you think you are? Personal home pages and self-presentation on the World Wide Web. Journalism & Mass Communication Quarterly. 1999; 76: 646-658.
27. Escalas JE and Bettman JR. Self-construal, reference groups, and brand meaning. Journal of consumer research. 2005; 32: 78-389.
28. Berger J and Heath C. Where consumers diverge from others: Identity signaling and product domains. Journal of consumer research. 2007; 34: 121-134.
29. Schembri S, Merrilees B and Kristiansen, S. Brand consumption and narrative of the self. Psychology & marketing. 2010; 27: 623-637.
30. Aaker J and Schmitt B. Culture-dependent assimilation and differentiation of the self: Preferences for consumption symbols in the United States and China. Journal of Cross-Cultural Psychology. 2001; 32: 561-576.
31. Gawronski B, Bodenhausen GV and Becker AP. I like it because I like myself: Associative self-anchoring and post-decisional change of implicit evaluations. J Exp Soc Psychol. 2007; 43: 221-232.
32. Morrison KR and Johnson CS. When what you have is who you are: Self-uncertainty leads individualists to see themselves in their possessions. Personality and Social Psychology Bulletin. 2011; 37: 639-651.
33. Kokkoris MD, Kühnen U and Yan S. Likes, dislikes, and the perception of choice as self-expression across cultures. Journal of Cognition and Culture. 2013; 13: 129-143.
34. Sharot T, Fleming SM, Yu X, Koster R and Dolan RJ. Is choice-induced preference change long lasting? Psychological science. 2012; 23: 1123-1129.
35. De Vries L, Peluso, AM, Romani S, Leeflang PS and Marcati A. Explaining consumer brand-related activities on social media: An investigation of the different roles of self-expression and socializing motivations. Computers in Human Behavior. 2017; 75: 272-282.
36. Byun KAK, Jones RP and Wooldridge BR. It is not always about brand: Design-driven consumers and their self-expression. Journal of Retailing and Consumer Services. 2018; 43: 296-303.
37. Aggarwal P and McGill AL. When brands seem human, do human’s act like brands? Automatic behavioral priming effects of brand anthropomorphism. Journal of consumer research. 2012; 39: 307-323.
38. Marques L, Borba C and Michael J. Grasping the Social Dimensions of Event Experiences: Introducing the Event Social Interaction Scale (ESIS). Event Management. 2021; 25.
39. Bearden WO and Netemeyer RG .Handbook of marketing scales: Multi-item measures for marketing and consumer behavior research.1999.
40. Dabholkar PA and Bagozzi RP.  An attitudinal model of technology-based self-service: moderating effects of consumer traits and situational factors. Journal of the academy of marketing science. 2002; 30: 184-201.