Investigation of Hong Kong Students ’ Esports Participation Intentions Using the Theory of Planned Behavior Approach: A Structural Equation Model

Esports have grown to become a core part of popular culture in many countries, including Hong Kong. Albeit the low participation rates in Esports in Hong Kong, it was starting to gain traction, yet the local Esports advocates were experiencing challenges in promoting and popularizing the Esport. Hence, the current study was aimed to identify and reveal the determinants of participating in Esports, as well as strengthen the work on Esports behaviors using the theory of planned behavior (TPB), a reliable and valid prominent theory in predicting human behavior across a plethora of contexts, ranging from health-related behaviors to sport consumption behaviors. In the current study, the convenience sampling method was used to recruit over 2000 students (secondary school, N = 1567 ( female = 615 ); university students, N = 1525 ( female = 255 ). The students were invited to participate in the survey for collecting their perception on Esports participation using TPB-based questionnaire. Results were analyzed using theoretical analysis and structural equation modeling (SEM). The results show that both secondary school and university students have positive attitudes toward Esports. The outcomes indicated that participating in Esports develops social networks, and people with professional Esports ’ skills capability and being fortitude tend to be models of Esports participation. However, inadequate resources are a signi ﬁ cant barrier to participation in the Esports business. The SEM model veri ﬁ ed that the variables of intention in Esports participation among the students in Hong Kong with an adequate goodness of ﬁ t index. As a whole, the current study has identi ﬁ ed the factors and determinants of Esports ’ intention and behavior among Hong Kong students, which were successfully displayed in terms of the theory of planned behavior. In addition, the ﬁ ndings are expected to provide the Hong Kong government with a documented framework to advocate Esports-related policies on a long-term basis.


Introduction
The last decade has seen a rapid rise of Esports around the globe. Once considered a fringe activity [1], Esports has become a core part of popular culture in many countries. Industry reports such as the one from Newzoo [2] substantiate this observation. The report revealed that in 2019, the number of global audiences for Esports was 443 million, and the industry made US$950 million. Even more staggering is the fact that these numbers are expected to grow even higher. The same report forecasted that by 2023, the Esports audience number would rise to 646 million, and the industry will become a US1.56 billion business.
However, since Esports has evolved from video game competitions in the 1950s [3], a negative reputation has always been attached. According to the Oxford dictionary (2018), video games are variously referred to "A game played by electronically manipulating images produced by a computer program on a monitor or other display," without competition system or prizes involved. Despite video games are considered as drivers of all Esports activities and competitions, Esports refers to "A multiplayer video game played competitively for spectators, typically by professional gamers." It requires skills, strategies, tactics, rules and regulations, institutionalized, teamwork, prize, and training, thus implies physical exertion and motivated performance by the internal and external awards [4]. Hence, its "sportification" nature makes it similar to traditional sports [5]. Yet, there were still different negative consequences of Esports which are well-documented in both the popular and research literature; some of the examples included adverse health impact [6,7], detrimental to academic pursuit, discrimination, and toxic online gaming culture [8,9].
Despite the questionable reputation, Esports is thriving in many places, especially among the youth population. In addition, more and more science educators are using video games as a means to engage students in inquiry-based science learning [10]. Under this backdrop, the sport marketing researchers are prompted to explore the motivators driving this current high level of engagement from the public, both as players and spectators. Subsequently, a wealth of research on this subject matter has been conducted. Nonetheless, there are still some places where Esports is only starting to gain traction, and the local Esports advocates are experiencing challenges in promoting and popularizing the sport.
Hong Kong, which has been slower than other Asian countries and regions in the development of Esports, did not have its first Esports tournament until 2016. Comparing with its neighboring regions and countries, the development of Esports in South Korea, Mainland China, and Singapore were originated in 1997, 2003, and 2008 [6,11,12]. There is a high level of interest and engagement in video games among Hong Kong youth. Studies have shown that video gaming is prevalent among local students. In a study that tracked the digital devices' usage of 7585 primary school students, it was found that 36.9% of them spent most of their screen time on video gaming [13]. Another study that involved 2006 primary school students also revealed that, on average, they spent 3.6 days and over 8 hours on gaming per week [14]. However, when it comes to adolescents aged 15-29 years, a survey showed that relatively less of them (13.4%) had participated in Esports competitions in the last 6 months [15].
Given this scenario, there is an apparent unalignment between the high interest of video gaming and the rate of Esports participation. To better understand, the grassroots information about their perceptions' of Esports participation in order to predict future Esports behaviour and to help practitioners in Esports developing areas design a comprehensive Esports marketing strategy. The present study aims to develop a model that can serve this purpose. This model's components are derived from the Esports Participation Questionnaire, an instrument developed by utilizing the theory of planned behavior (TPB) [16]. The three dimensions that comprise the TPB-attitude toward the behavior, subjective norms, and perceived behavior control provided the theoretical platform to delineate individuals' perception of Esports and demonstrate how that perception impacts the intention and behavior toward Esports.
This study will also further extend the literature on Esports behaviors. The research team argues that the benefit of employing TPB to examine Esports behaviors is twofold. First, this approach reveals the factors that influence the intention to engage in the activity. Second, it also allows the researchers to validate if there is a disconnection between the intention to participate in Esports and the actual Esports behavior, which causes the low participation rate of Esports in Hong Kong. At the field level, the findings from this research should afford the industry practitioners with information to design an effective strategy to educate and market Esports to the public. It will be crucial in building a core fanbase to further develop and substantiate the Esports industry locally.

Literature Review
The rapid growth of Esports and its economic development potential had been noted in Hong Kong, where Hong Kong Esports (HKE) Limited was established in 2013. Since then, it has been hosting the Hong Kong Esports Finals annually [17]. Subsequently, several other Esports organizations were also established, namely, Hong Kong Student Esports Association in 2013 and "E-sports Association Hong Kong" (ESAHK) in 2015. In 2017, the government funded HK$35 million for the first Hong Kong Esports Festival, which attracted around 50,000 video games, K-pop fans, and tourists to Hong Kong. In the 2018-2019 budget, the government allocated HK$100 million to Cyberport to promote the Esports industry's development.
While the government in Hong Kong has started providing financial resources to jumpstart the industry, the development of Esports is still considered at the infancy stage. In comparison, Esports amass to substantial economic returns in other countries. According to an analysis of individual markets from Newzoo [18], other significant markets besides China include the US (with 186 million players, 48.5 million Esports audience, and generated US$32.9 billion revenue), Japan (with 66.8 million players, 12.4 million audiences, and US$17.9 billion in revenue), and South Korea (with 28.4 million players, 12.1 million audiences, and US$5.76 billion in revenue). Moreover, it should be noted that even though China has the most significant market with the sheer volume of players and revenue generated, South Korea is considered by many the birthplace of professional Esports [19][20][21]. Esports has been developed for a much longer time in Korea and is entrenched into the mainstream culture [21]. It is just as typical for a group of people going to the internet café called PC bang locally to play Esports as going to the movie theatres on the weekends.
Given the Esports development level in our neighboring countries, it is evident that Hong Kong has a lot of room to grow in Esports. Yet amid the tremendous economic potential Esports can offer, the current Esports participation rate in Hong Kong is relatively low. One of the reasons can attribute to the public's negative attitude toward Esports/gaming. For the most part, they are concerned with the impact of Esports participation and long screen time on mental and 2 Human Behavior and Emerging Technologies physical health, such as gaming addiction or risky behaviors [22,23]. Previous local research has highlighted such concerns. In 2014 and 2017, the Department of Health, Hong Kong (DH), surveyed the healthy use of the Internet and electronic screen products among students from 24 local secondary schools across the territory. The survey found that 33.9% of secondary school students spent more than 3 hours a day on average on internet and electronic screen products. Furthermore, over 70% of parents reported that they had arguments with their children about their screen time. Although the public also expressed concerns about video games, such as "gaming addiction," however, there is currently little clarity that has been achieved in terms of diagnosing the criteria and appropriate symptoms of "gaming addiction," as "video game addiction" is not a stable concept and clinical impairment may be low; given the ongoing controversy, the classification of video game addiction as a mental disorder is premature [24].
Instead of predicting the negative uncertainties of the Esports industry, it is better to embrace the industry as a new zeitgeist and harness its features to bring potential benefits it brings to society. More recently, increasing research has shown the benefits of playing active video games, specifically physical and psychological health [25][26][27]. Merino-Campos and colleagues [28] conducted a systematic review on benefits in skills and health, and the results found that active video games could improve motor, intellectual, and physical capacities related to physical activity and education, such as increased moderate-vigorous physical activity. These positive outcomes were also established in Chinese studies [27,29]. Other technical skills including improvement in visual-spatial abilities, hand-eye coordination, and action speed were also suggested to be connected to the practice of Esports [1,[30][31][32]. To further demonstrate the physical aspect of Esports participants, Kari, Siutila, and Karhulahti [33] conducted a three-year mixed quantitative and qualitative studies among 115 professional Esports players. The results showed 55.6% of them did physical exercises to contribute to their Esports career, and on average, they spent 1.08 hours daily on exercising. In terms of Esports player's mental health, lots of researchers found that Esports helped to improve self-regulatory skills, as well as cognitive, motivational, emotional, social benefits, and life goal achievement [1,31,34,35].
As there is a lack of consensus on the impact of Esports/ gaming currently, it is reasonable to believe that there is room for Esports advocates to change the public's perception of the sport. Moreover, for a developing market like Hong Kong, the field practitioners must have grassroots information that can help them to devise a comprehensive marketing strategy. For these reasons, the current study aims to access the people's view on Esports participation in Hong Kong. Specifically, the study will focus on the local youth population. According to Cyberport [17], most of the population playing Esports in Hong Kong were age 15 to 21 years old (85%). Developing the Esports industry to keep pace with the world will help promote the Hong Kong economy and create more divertive career opportunities for local young people. In line with this thinking, the current study aims to (a) examine the participation of Esports among secondary and university students in Hong Kong; (b) use theory-based determinants (the attitude, perceived norms, and perceived behavioral control) to explore the participants' perception toward Esports among secondary and university students in Hong Kong; and (c) to examine the relationship among the determinants using the theory of planned behavior model.

The Theory of Planned Behavior and Research in Esports
Behaviors. TBP has been a prominent theory in predicting human behavior across a plethora of contexts, ranging from health-related behaviors [36,37], volunteer behavior [38,39], physical activity and exercise [40,41] to sport consumption behaviors [42][43][44]. The theory was extended from Ajzen and Fishbein's previous work on the theory of reasoned action (TRA) [45,46]. TRA was evolved from Fishbein's [47] work on the relationship between attitude and behavior. Under this premise, the attitude refers to the individual's evaluation of the behavior and the expected outcome for undertaking said behavior [48]. The gist of TRA is that the proximal determinant of volitional behavior is the behavioral intention, and the intention embodies an individual's conscious decision to put in the effort to execute the said behavior [49]. In TRA (and later TPB), behavioral intention is influenced mainly by an individual's attitude toward the behavior; thus, it acts as a mediator between attitude and behavior [50].
Another determinant in the TRA is termed as subjective norm. It depicts the social pressure individual encounters from the significant people or groups to undertake (or not undertake) the said behavior. Critically, TRA was posited within the perimeter where the said behavior was volitional. Therefore, as Fishbein [51] alluded to later, the predictability of this TRA model was not as efficient when (1) the said behavior was not wholly volitional, (2) there was a choice problem involved in the scenario, and (3) the individual's intention was measured when insufficient information was available to formulate a confident intention ( [52], p. 325). A third component, perceived behavioral control (PBC), was introduced to the TRA model to rectify this limitation. As a result, TBP was formed.
PBC concerns situations when the behaviors are not completely volitional, when the individual's perception of control over the said behavior would affect their intention toward the behavior, thus contributing to the prediction of behavior. In essence, this dimension entails "the perceived ease or difficulty of performing the behavior" ( [16], p. 188). Furthermore, it is also related to the individual's perception if they have the necessary skill and resources to undertake the behavior and the potential obstacles presented in relation to the execution of the behavior ( [53], p. 474). Essentially, TPB has remedied the shortcoming of TRA by enhancing the applicability of this original theory to those more complex behaviors that are not necessarily volitional. In line with this view, the TPB model looks pertinent in the current research context, especially when Esports is developing in Hong Kong. The perceived lack of skills from 3 Human Behavior and Emerging Technologies the players and the shortage of opportunities to participate regularly can be considered obstacles that influence the PBC. Consequently, it can impact the behavior intention toward Esports as well.
The research of Esports behavior thus far has been mostly revolved around the motivations for consumption (e.g., [32,[54][55][56][57]), and only a few have incorporated a predictive element into their studies (e.g., [58,59]). Therefore, with the utilization of TPB, this study intends to continue this line of research and to further expand the work on Esports behaviors through an established behavioral theory. This study will also look to provide the industry practitioners with information that would aid them in predicting consumption behaviors from the potential clients.

Method
3.1. Participants. Based on the survey conducted by Cyberport [17] and Hong Kong Federation of Youth Groups [15], the major population of Esport players was aged 15-29 years old. Hence, the population of this study was Hong Kong secondary school and university students aged 29 years or below. Other inclusion criteria for participants are permanent residency in Hong Kong and literacy in Cantonese.
In the current study, a total sample size of 2115 students has participated in the survey, which is an adequate sample size to an estimated parameter (at least 10 : 1 ratio) for confirmatory factor analysis and structural equation modeling as suggested by Hu and Bentler [60]. Also, this samplesized has met the guidelines of conducting a populationbased survey suggested by the World Health Organization [61], which is within 2% of the true population of Esports participation (13.4%) reported by Hong Kong Federation of Youth Groups [15].

3.2.
Procedures. The period of this study was from March to September 2020. At first, the newly developed questionnaire had gone through a pilot test with five university students (three of them were participants and two nonparticipants) and two Esports experts from ER Esports to evaluate the questionnaire's length, wording, and formatting. Before data collection, an invitation letter was sent to randomly selected secondary schools. Both paper and web-based questionnaires with consent forms were provided to the schools as options. Due to the COVID-19, there was no positive reply from the selected schools, so convenience sampling was applied to reach other schools in the districts. With principals' consent, a trained research assistant sent web-based questionnaires or handed out paper questionnaires for data collection to the responsible teachers.
An invitation e-poster attached with a QR code for the web-based questionnaire was sent to students via mass email for the universities. In schools where we could not obtain the student mailing lists, the e-poster was sent to the student affairs office or physical education teachers at the school for promotion. All participants completed only one time, and the average completion time was with 10 minutes. Prior to administering the questionnaires, permission was granted from the institutional review board for research with human subjects of the university. Informed consent forms accompanying the questionnaires were sent to secondary schools for approval by the principals and parents prior to recruitment of students under the age of 18. Similarly, an informed consent form was attached to the web-based questionnaire, and before completing the questionnaire, participants were asked to tick a box to clearly read the study and sign their consent to participate in this study. After collecting data, all participants were entered into our random drawing, and 50 winning students received HK$500 supermarket vouchers. Participants were informed of the confidentiality of the data and free to discontinue their participation in this study.

Measures.
Participants were asked to respond to questions about their demographic information, including their age, gender, education level, school, housing, and family income. Next, based on the theory of planned behavior (TPB), Esports Participation Questionnaire was developed to explore Hong Kong teenagers and young adults' Esports participation and their respective attitude toward the sport.
The Esports Participation Questionnaire [62] uses standardized questions, tapping into TPB constructs to measure participants' attitudes, subjective norms, and perceived behavioral control (direct measures), and indirect measures as suggested by Ajzen [16]. The questionnaire was on a 5point Likert scale. The items of the direct measures included the following: (1) Intention to participate in Esports in the coming six months (4 items) Example question: "In the coming six months, I will participate in Esports competitions" 1 (strongly disagree) to 5 (strongly agree)

(2) Attitude toward Esports participation (4 items)
Example question: "Do you think your participation in Esports competition in the next six months will be?" 1 (very harmful) to 5 (very beneficial)

(3) Subjective norms about participating in Esports (4 items)
Example question: "Many people or groups I value think I should be competing in Esports in the next six months"1 (strongly disagree) to 5 (strongly agree) (4) Perceived behavioral control of participating in Esports (4 items) Example question: "It will be easy for me to compete in Esports in the next six months" 1 (strongly disagree) to 5 (strongly agree) Apart from the standardized questions for direct measures, the indirect measures' items were developed based 4 Human Behavior and Emerging Technologies on the results of the qualitative study [63]. The interview guide of this qualitative study was developed using TPB; the results revealed the perceptions of Hong Kong younger adults on the beneficial and deleterious outcome, subjective norms, and barriers and facilitators of Esports participation. According to the above outcome, a residual section on the questionnaire, the indirect measure, was identified following Ajzen's [64] questionnaire structure to examine more Hong Kong teenagers' behavioral, normative, and control beliefs on Esports' participants. For behavioral beliefs, items are formulated to assess the strength of the behavioral beliefs and the evaluation of the outcome or experience. It includes Hong Kong teenagers' views on the beneficial and deleterious outcomes (goal setting and achievement, physical health, socialization and teamwork, psychological benefits, academic and time distribution, physical strain, negative social image, and perception toward Esports participation). The example questions were below: (1) Benefits outcome: increase social skills Behavioral belief strength "My participation in Esports in the next six months will increase my social skills (meeting new people, improving friendships)," 1 (strongly disagree) to 5 (strongly agree) Outcome evaluation "Increase my social skills (meeting new people, improving friendships) is…" 1(extremely bad) to 5(extremely good) (2) Deleterious outcome: time-consuming Behavioral belief strength "My participation in Esports competitions in the next six months will be time-consuming)," 1 (strongly disagree) to 5 (strongly agree) Outcome evaluation "The time-consuming for me is…" 1 (extremely bad) to 5 (extremely good) Normative beliefs are formulated to assess the strength of the normative belief and the motivation to comply with the referent individual or group. It includes participants' perceived social norms (people with particular characteristics or conditions as modeling) or significant others (parents, peers, teachers). For example: (1) Significant others: parents Injunctive normative belief strength "Do my parents and family think I should compete in Esports in the next six months?" 1 (strongly disagree) to 5 (strongly agree) Motivation to comply "I want to do what my parents and family think I should do", 1 (strongly disagree) to 5 (strongly agree) Control beliefs are formulated to assess the likelihood that the factor will be present and the factor's power to facilitate or impede performance of the behavior. It is about the barriers and facilitators (balance between academic and Esports, capability, career prospect and future reality, psy-chological benefits, and peer encouragement and support) of participating in Esports. For example: (1) Sample control factor: academic hindrance Control belief strength "I expect my academic to be hindered by my participation in Esports in the next six months." 1 (strongly disagree) to 5 (strongly agree) Power of control factor "The academic hindrance will make it easy for me not to compete in Esports for the next six months." 1 (strongly disagree) to 5 (strongly agree) According to Ajzen [65], the equation of indirect measures is summing the products of control belief strength times perceived power, for example, if participant perceived the belief strength of participating Esports in the next six month will increase he/she social skills is 4, and the perceived power of increase my social skills is 5, then the total perceive behavioral control of social skills is 4 times 5, which is 20, hence, the maximum rating of the indirect factors will be 5 times 5 is equal to 25.
The reliability and validity of the Esports Participation Questionnaire were documented as well. All scales involved in the questionnaire showed adequate internal consistency with Cronbach's α above 0.8. It showed adequate goodness of fit indices in all subscales, with all subscales' chi-square values below 5, comparative fit index (CFI) and nonnormed fit index (NNFI) rated 0.90 or above, and standardized root mean square residual (SRMR) and root mean square error of approximation (RMSEA) values below.08 [66]. Because the confirmatory factor analysis showed some different outcomes among the secondary school students and university students, some different outcomes were deleted correspondingly to access the goodness of fit model; however, to show more determinants of Esports intention of participants, this study separate two samples (secondary school and university students) for analysis (please see Figure 1 for the investigation model).

Statistical
Analyses. The data were analyzed using Statistical Package for the Social Sciences (SPSS) (Version 26). Descriptive statistics, such as frequency, mean, and standard deviation, of the demographic information and the TPB variables or factor were computed to examine the prevalence of Esports participation and the factors influencing Esports participation among Hong Kong secondary school and university students. An independent t-test was done to identify potential differences in the influencing factors between Esports players and general secondary school and university students, as well as between gender. The TPB model factors were also computed to examine the prevalence of Esports participation and the factors influencing Esports participation among Hong Kong university students. Pearson correlation was used to analyze the relationship among the influencing factors among participants' intention and behavior in playing Esports. The confirmatory factor analysis (CFA) and structural equation modeling (SEM) were conducted by using the LISREL (Version 10.3). Under the usage of the maximum likelihood estimation (MLE), the goodness of fit models were determined according to the following criteria: (1) chi-square ranged 2 to 5, (2) comparative fit index (CFI) [60] and non-normed fit index (NNFI) rated as .90 or above as a model of good fit [60], (3) standardized root mean square residual (SRMR) value of .08 or below as a model of good fit [60], and (4) root mean square error of approximation (RMSEA) value as .08 or below, with a 90% confidence interval that holds within this value which was considered as a model of good fit [66].

4.1.
Demographics. In the current study, twenty-five secondary schools have agreed to participate in this study. These schools were located in 11 Hong Kong districts (New Territories and Hong Kong Island). A total of 1576 students have completed the web-based and hardcopy questionnaires. After data clearance that excluded the missing and outliers, 1164 (male = 533; female = 615) valid data were included in the study. Among the secondary school students, fortyeight of them had participated in Esports in the last six months. Because according to the psychometric properties measurement of TPB mode [62], good fit when separated secondary school and university samples. Therefore, in this study, the two samples of secondary schools and universities were analyzed separately.
On the other hand, ten UGC-funded universities with 1525 web-based questionnaires had been returned. After data screening and cleaning of cases from incomplete questionnaires, 951 university students went into further data analysis. Among the 951 students (male = 490; female = 255), 56 participated in Esports in the past half years. Most of them were degree students (60.5%) studying in Kowloon City (78.2%), and nearly 27.3% of students reported that their household income was USD 3,850 or higher. Participants' demographics are shown in Table 1. For the university student, a vast majority of them (94.1%) also stated they did not participate in any Esports competition in that period, with the remaining 5.9% said they did. For those university students (n = 56) that had Esports experience, on average, they participated 14. Regarding their favorite games, the two favorite games among the secondary students who competed in Esports were PUBG (45.8%) and Rainbow 6 (39.6%), while PUBG (41.1%) and League of Legends (LOL) (39.3%) were the preferred game of choice among the university Esports competition participants. Most of the respondents' parents (62.5%) in secondary group knew about the Esports participation and showed support (68.8%). And 78.6% of the parents in university group knew the respondents competed in Esports; among these respondents, 62.5% feel their parents supported their involvement. Esports participants are shown in Table 2.   and inadequate resources to support Hong Kong Esports development (M = 12:92, SD = 5:73) were deemed to be the most significant barriers to partake in an Esports career, regardless of gender difference.

Esports Competition
As for the university student, the results also indicated that inducing family conflicts due to Esports gaming (M = 5:63, SD = 3:24) was the least influential factor in affecting the students' perceived behavioral beliefs in Esports participation. Instead, they felt that gaining social networks  Table 3.

Differences in Esports Participation Intentions and Other TPB Direct Measures in terms of Esports Participation and
Genders. The results suggested there were significant differences in all the TPB direct measures. The construct of Esports participation intention showed the largest mean difference at M = 1:41, SD = 0:3; F ð1, 1090Þ = 89:75, p = :000. Because of the substantial differences in variances between Esports participants and nonparticipants group, the Welch's test was conducted. The results from the Welch's test also indicted a significant difference between the groups F W ð1, 1090Þ = 112:52, p = :000. As far as the comparison between the male and female groups, an independent t-test was conducted and found that there were significant differences in the intention to participate, attitude toward Esports, subjective norms, and perceived behavioral control (males are more likely than girls to be intention in participating in Esports, to be influenced by people or groups who value them, and to perceive that they are more likely to participate in Esports). In contrast, there was no significant difference in attitudes toward Esports, tð734Þ = 1:85, p = :065.
Similar to the secondary student group, the result from university student group also indicated significant differences in all TPB direct measures between Esports competition participants and non-Esports competition participants, with the level of intention in Esports participation that showed the largest mean difference at M = 1:47, SD = 0:12; Fð1,949Þ = 144:07, p = :000. Because of the substantial differences in variances between Esports participants and nonparticipants group, the Welch's test was conducted. The results from the Welch's test also indicted a significant difference between the groups F W ð1,949Þ = 105:04, p = :000. Moreover, there were significant differences in the intention, Additionally, Esports competition participants rated comparatively high in perceiving physical injuries as a negative aspect of participating in Esports, in which this was not highly rated among non-Esports competition participants.  Table 4 illustrates the correlation matrix between the dependent variables and independent variables (intention and behavior) among Hong Kong secondary school and university students. The correlation results indicated that the independent variables, including attitude, subjective norms, perceived behavioral control, positive factors, negative factors, significant others, personal conditions, and barriers, were significantly correlated with the intention to participate in Esports. Conversely, university students indicated that Esports' negative factors and policy had a significant negative correlation with both their intention and behavior. Furthermore, despite Esports' negative factors and policy had shown to affect secondary school students' intentions, it did not show significant association with their behavior in participating in Esports.
Before conducting the structural equation modeling (SEM), confirmatory factor analysis (CFA) was done to affirm the goodness of fit of each measurement model. Table 5 shows the results of the CFA for all the individual measurement models of the current study. All measurement models achieved a good fit model upon conducting SEM.

Structural Equation Model
(SEM) Results. The structural equation model was formed according to the TPB model, with "positive" and "negative" as the latent indicators of behavioral beliefs; "significant others" and "personal condition" as latent indicators of normative beliefs; and "barriers," "facilitators," and "policy" as latent indicators of control Beliefs. They were linked to the respective direct measures, including attitude, subjective norms, and perceived behavioral control, respectively. See Figure 1 for the examined TPB model. Table 6 shows the results of the SEMs' goodness of fit index of both target groups, and Table 7 shows the summary of goodness of fit of the measurement models. . In other words, the insignificant path did not significantly affect the overall model fit in representing the intention aspects of Hong Kong Secondary School students in participating in Esports. The path result of the model diagram is shown in Figure 2 and Table 8. 4.9. University Students. Among the university school students sample, the proposed model showed an adequate model of fit indices with X 2 ð4236:57/1264 = 3:37Þ, CFI = 0:906, NNFI = 0:87, SRMR = 0:094, and RMSEA = 0:056 (90%CI = 0:054 − 0:058). In addition to the original CFA results, three items from behavioral beliefs and one item from control belief were removed due to the high standardized residual (32.2-50.2). Moreover, other than the covariance identified in the CFA results and the path originated from the TPB model, additional correlations between items in the direct measures were found during the SEM analysis. All direct and indirect measures, including positive factors, significant others, facilitators, and barriers of Esports, showed significant pathways toward Esports intentions. Yet, the model has also shown insignificant standardized solutions between "negative" components of behavioral beliefs and attitude, between "personal condition" components of control beliefs and subjective norms, and between the "policy" component of control beliefs and perceived behavior control. With the particular path removed, the model fit indices showed insignificant change in the chisquare value by just increasing for 0.02 (X 2 (4294:36/1267 = 3:39). In other words, the insignificant paths did not significantly affect the overall model fit in representing the intention aspects of Hong Kong university students participating in Esports. The path result of the model diagram is shown in Figure 3 and Table 9.

Discussion
The current study aims to investigate the Esports participation behavior and intentions among Hong Kong secondary school and university students, using the theory of planned behavior approach. This study suggests that the development of Esports in Hong Kong is still in the early stage, as reflected in the low participation rate in the Esports competition. However, it should be noted that since recent reports have shown there are still more male Esports gamers than female [67,68] , therefore, this study may be slightly biased toward nongamer, as it has surveyed more females (52.6%, among those that disclosed their gender). The outcomes of the SEM models have summarized the variables that could have affected Hong Kong secondary school students' and university students' intention in Esports participation. Among the secondary school students, all the variables showed significant association with their intentions in participating in Esports. Yet, the path of the negative factors of Esports toward attitude was shown as insignificant in the SEM model. Most of the variables, except the facilitators and policy, showed significant association with their intention among the university students. Yet, the path of negative factors of Esports and the personal condition of playing Esports, as well as policies related to Esports, toward the 10 Human Behavior and Emerging Technologies respective direct measures was insignificant in the SEM model. It should be noted that the Esports behavior of students was not included in the SEM model analysis due to the inconsistency of the type of data; the SEM model was not able to reveal outputs of data that was combined with continuous and categorical data. When closely examined the indirect measures of the TPB model, both the secondary and university student  11 Human Behavior and Emerging Technologies groups revealed similar factors that deem to have significant impacts on Esports participation. For instances, the similar sentiment shared by both groups on attitude toward Esports competition suggests that they are drawn to Esports by the social elements. This finding is also similar to the previous studies on the more casual gamers [69,70]. Communication skills development and achieving team works are also considered important as well, akin to the findings on video gaming in earlier literature [71][72][73][74]. In terms of the factors that impact Esports participation, both groups expressed skill level, and fortitude play a significant role. Also, while the research team assumed family and friends would be influential in the respondent's participation, given the study's sample mostly from a Chinese background, the result is shown otherwise. It can be implied that most young people are so used to having family and friends disapproved of their video game hobby. As a result, they do not base their participation decision on this negative feedback anymore. Finally, both cohorts perceived the lack of skill and inadequate resources from the local Esports industry as the two major barriers    12 Human Behavior and Emerging Technologies that prevent them from engaging Esports at a higher level. The local Esports marketers should be able to utilize this information when addressing the preconception about the skill level needed to participate in Esports, as well as the need to highlight the social aspects of Esports to the potential consumers. These findings reflect currently the local Esports participants that do not necessarily consider Esports as a competitive endeavor or a potential career path. Instead, they deem Esports as a progression from video games, with the massive online battle arena (MOBA) games like PUBG and LOL that provide platforms for them to socialize, develop communication skills, and experience teamwork success. Collectively, the findings signified in this early development stage are that it is more sensible to market and highlight the social aspects of Esports so that it can attract more interest and participation initially. It is critical to nurture the local gamers' interest in Esports so that a group of core fans and competitors can be formed where the industry to be built upon. As the indus-try becomes more mature and amass a stable group of followers, it could be beneficial to emulate the model in South Korea, where a government-sanctioned Esports association was established with the aim to promote and develop Esports as a viable career path [75].
Previous literature had identified various significant factors that would have affected people's Esports' consumption and intentions [32,57,76]. Despite Sjöblom and Hamari [57] research only revealed the factors that affect people's intention to watch Esports, the mentioned factors like social integration, personal integrative, and releasing tension were shown as significant positive factors in affecting Esports' participation in the current study. Moreover, research studies have shown determinants, including hedonic motivation, effort expectancy [58], and gaining self-concept [77], as critical factors in affecting Esports' intention, in which these determinants were also in line with the positive determinants stated in the current study, such as enhancing students' perseverance in achieving the goal of Esports, as well   [32,78]. At the same time, this is also related to the identified normative beliefs latent factors-significant others. However, with most of the existing literature focused on Esports players, it might not identify other significant others as the current study did, such as teachers, parents, and the general public.
Other than the positive factors and significant others, the facilitators and barriers of participating in Esports were significant in affecting Esports perceived behavioral control as to intention. The research of Lee and Schoenstedt [32] revealed that having a competition to show and display one's Esports ability and equipped skills for games were significant impacting factors of Esports game playing, thus reflecting the identified facilitators of the current study. Besides, a systematic review [77] revealed that mastery of skills, pursuing self-improvement, gaining self-esteem, accomplishment and social recognition, and being equipped with formal institutional rules and Esports governance were essential elements that attracted players to treat Esports as a career. These elements were considered a holistic cover of the facilitators and barriers indicated in the current study, yet even in conformity with some of the identified policy indicators. Although the policy determinants were revealed as insignificant in affecting Hong Kong university students' Esports intention, it is worthy of being indicated in general, which could also facilitate new policy initiatives and implications for future investigation.
Additionally, secondary school students were also affected by the personal condition of being an Esports player. In the same systematic review [77], it was identified that the characteristics of Esports players include (1) being with great knowledge and skills in video gaming, (2) able to make fast and strategic decisions, (3) enabling self-motivation, (4) enable a game-life balance either during competition or training, (5) able to cope with harassment, (6) able to maintain a positive mindset, and (7) with a stable physical and mental condition. It is worth noting that the current research had indicated up to 5 of the elements mentioned   14 Human Behavior and Emerging Technologies above. Essentially, Hong Kong students showed a common perception toward the traits and quality of being Esports players, and the secondary school students, in particular, treated these traits as essential as well.
It is worthy to note that the result of this study differs from previous related literature. For example, utilizing the uses and gratification theory, Weiss and Schiele [79] found that competition, challenge, and escapism prompt continuous Esports gameplay, but not social relationship. Likewise, in recent research that grounded in the Unified Theory of Acceptance and Use of Technology 2 [80], Jang and Byon [58] discovered four critical factors, hedonic motivation, price value, effort expectancy, and flow, to be influencing in Esports consumer gameplay intention. With the research subjects from these two studies being more experienced Esports players, it is reasonable to presume that as the players' career progress, their motives to engage in the sport will change as well.
Finally, another area of interest from the findings is the difference in participation intention and TPB measures between genders. The findings indicate that there are difference between the two gender groups across both secondary and university students. There were significant differences in the intention, attitude, subjective norms, and perceived behavior control over Esports in the secondary students group; there was significant differences in the intention, subjective norms, and perceived behavior control, but no significant difference in the attitude in the university group. This finding entails that potentially different psychological factors will impact intention to participate in Esports unequally between the two genders. More research is needed in a different context to validate this finding. However, for the practitioners in the field, this finding signified that it would be advantageous to focus on these psychological factors when devising specific marketing plans for the different two gender groups in both secondary and university school students. Indeed, the current substantial differences in participation levels between genders can allude to a homogenized marketing strategy without a specialized promotional plan toward the female demographic. An industry-wide marketing campaign that focuses on educating the public about contemporary gaming and removing participation barriers might yield a more effective result in recruiting new players, especially the female Esports players.

Contribution
As a whole, the current study has identified the factors and determinants of Esports' intention and behavior among Hong Kong students, which were successfully displayed in terms of the theory of planned behaviour. Although the goodness of fit index was not as perfect as assumed, it provided an adequate picture in reflecting the overall interrelationship between the determinants and intentions of Esports. Hence, it could be seen as able to contribute to the literature on Esports and its related industry by identifying the determinants of students' participation in Esports, including the possible health issues created and the mental status achieved, as well as identifying the relating governing structure of Esports in terms of a theoretical framework. The following paragraph will discuss more potential contribution suggestions on Home Affairs Bureaus and Leisure and Culture Services Department, Education Bureau, and Hong Kong Schools Sports Federation.
Firstly, to bring local Esports to be in line with the world, more regular Esports competition and tournaments should be hold and create employment opportunities for potential Esports professionals, and Esports athletes, such as coaches, referred, anchors, or tournament managers to ensure that their Esports journey would not be interrupted.
Furthermore, the current research being part of a Hong Kong public policy research on "The Development of Esports in Hong Kong" also aimed to provide the Hong Kong government with a documented framework to advocate Esports-related policies on a long-term basis; Education Bureau is suggested to provide guidelines on developing Esports-related education courses and programs in the current education curricula, as well as providing Esports practitioners or game companies with a cooperation guide for developing Hong Kong's Esports markets; actions include standardizing the teaching and practice of Esports, working with educational institutions to build a healthy Esports environment, developing Esports-related expertise (e.g., video content creation, event planning and operations, event commentary and hosting, and Esports club management), and proving the legitimacy of the industry and the viability of Esports education.
In addition, investing funds to cultivate professional Esports players, establishing an Esports association to support Esports, provide expertise and advice, and increasing the visibility of Hong Kong Esports in the international area. For example, to enhance Esports development in secondary schools and universities, the government can consider providing financial support for setting up training facilities as well as forming Esports school teams to nurture future Esports talents and offer interschool Esports competition opportunities. For the effectiveness of these policies on public, including parents and teachers, it should be discussed and investigated in future research.

Data Availability
According to the approval ethic of informed consent statement: The answers participants provide in the study are confidential, which means that only the study investigators will have access to any information they share. All the data will also be kept confidential in any publication of the results of this study. The data will be reported/written in aggregate terms. The material will be maintained for up to 3 years.

Additional Points
Highlights. (i) Both secondary school and university students have positive attitudes toward Esports. (ii) Inadequate recourses in supporting Hong Kong Esports development is a significant barrier for both university and secondary school students to participate in the Esports business. (iii) The TPB model was validated as a potential model to study