How Digital Skills Influence on Digital Participation in China? The Mediating Roles of Online Interpersonal Communication and Online Immersion

This study examined the influencing factors of digital participation, including digital skills, interpersonal communication, online immersion and mediating roles of online interpersonal communication and online immersion. This study selected 1,204 samples from the 2017 Chinese General Social Survey. Amos 26.0 was used to test the theoretical model using the structural equation model (SEM). The study found that digital skills positively affect digital participation, digital skills positively affect online interpersonal communication and online immersion, and online interpersonal communication and online immersion positively affect digital participation. Online interpersonal communication and online immersion were found to significantly mediate the nexus between digital skills and digital participation. Moreover, income was found to be no longer a significant factor in the relationship of digital skills on digital participation. This study found the mediating roles of online interpersonal communication and online immersion between digital skills and digital participation, which was not available in previous studies. This result proved that the interaction between participants and the subjective immersive experience via the network are crucial for digital participation, and that the close information communication with different populations and an immersive network environment can facilitate the interactive expression of participants in cyberspace. The study also found that at least in China, the income gap is no longer a major factor affecting the relationship between digital skills and digital participation. Plain Language Summary The purpose of the study is to examine the influencing factors of digital participation, including digital skills, interpersonal communication, online immersion and mediating roles of online interpersonal communication and online immersion. The study selected 1,204 samples from the 2017 China General Social Survey. Amos 26.0 was used to test the theoretical model using the structural equation model (SEM). The conclusions are that this study revealed the key role of digital skills in improving digital participation and proposed strategies to strengthen digital skills training and education and optimize digital platform page operation. In addition, this study examined the positive effects of online interpersonal communication and online immersion on digital participation. The implications are that this study demonstrated that online interpersonal communication and immersion played an intermediary role in the process from digital skills to digital participation, which enriched the understanding of the changes of communication relationship and physical and mental perception produced by individuals through networks. Meanwhile, this study indicated that gender, education, and regional differences influenced the results, and it is necessary to further improve the measures of digital platforms in content guidance and target group selection. The limitations of the study are that this study was mainly aimed at adults, as data on the digital participation of adolescents is insufficient. Meanwhile, this study did not investigate a specific group, and future research could be targeted from the economic, education, and other aspects of classification research.


Introduction
By the end of 2021, the number of Internet users in China had reached 1.032 billion, and the Internet penetration rate was 73% (CNNIC49, 2022).The amount of Internet access in China has expanded rapidly in recent years, with the advantages of network development, diversity, and real-time communication affording individuals' broad opportunities to fully integrate into society and observe social developments.Scholars have classified digital skills and discussed the participation tendency of different social groups (Hargittai & Jennrich, 2016;Scheerder et al., 2017).Inefficient digital participation not only heightens the loss of discourse in cyberspace but also aggravates the emergence of social inequality (Ragnedda et al., 2022).Based on the theory of planned behavior (TPB), there is a positive relationship between attitude and online campaigning (Chen et al., 2019), and that subject norms are a significant determinant of intention across all the age groups (Oni et al., 2017).Furthermore, trust in others is a key driver of behaviors in technology usage (Tams et al., 2018).Media platforms that pay attention to user experience promote the formation of online communities.Online communities include interactions between people in cyberspace who share the same interests and beliefs (Bimber, 2014).Reciprocity is the perception of supportiveness between the members of a participatory community (Adams, 1999).
Digital participation includes individual behaviors and close contact with other participants.Such connection occurs through online communication on social media, which can bring closer intimacy with others and promote individual digital participation.Extant research has focused on the inequality caused by digital skills and participation, ignoring the impact of long-term online immersion on these factors.Individual psychological willingness to participate rather than participation ability merits investigation.In addition, the current study lacks the digital skills and participation of developing countries from the perspective of a larger sample.This study used national survey data from the Chinese General Social Survey (CGSS) to examine influencing factors of digital participation and digital skills among people in China, construct a model framework, and analyze the impact of individual gender, age, income, education, regional differences on digital skills and digital participation with the goal of promoting digital participation.
The remainder of this paper is structured as follows: The literature review provides theoretical support and research hypotheses for the main factors affecting digital participation.The following chapters describe the proposed research methods and data analysis results.Finally, the practical significance of influencing the results is discussed, and the limitations of the research and the future research direction are explained.Figure 1 presents a chart of the framework of the study.

Literature Review
Digital participation is a type of behavior, where users communicate and interact on a network platform after connecting to the Internet through information equipment.The development of information technology has become inevitable and governments have begun to promote public participation in digital information technology.The rise of information and communication technology (ICT) has created the possibility of digital participation and plays the role of a communicative bridge between individuals in a society (De Wit et al., 2012).Furthermore, digital participation provides a brand-new environment for the expression of opinions, offering the immediacy and full coverage of the network, allowing service interaction and enabling people to access peer opinions (King & Xia, 1997).Digital participation helps the public better intervene in cyberspace, promotes communication, and overcomes existing social divisions and inequalities (Pham & Massey, 2018).However, the existence of a digital divide creates challenges for digital participation.Geographic location causes unequal access and contributes to differences in ICT access levels (Sylvester & McGlynn, 2010).Network access is easier in urban than rural areas, and rural areas have many shortcomings in the universality, effectiveness, and scientificity of participation.Children have traditionally been excluded from debates about citizenship, as they are often considered not-yet or future citizens (Barassi, 2019).However, today's children are digitized before birth and participate in society digitally; therefore, children and adolescents should be considered in digital participation (Literat et al., 2018).In addition, the differences caused by economic inequality, including social stratum (Pearce & Rice, 2017) and degree of educational literacy (Joseph et al., 2012), may lead to insufficient digital participation.
Researchers have explored how to improve digital participation and promote people's willingness to participate.Smorgunov and Ignatjeva (2021) demonstrated that the use of digital platforms was determined to a greater extent by affordability than by efficiency and security, and trust in institutions was more influenced by efficiency and security.Scholars have linked digital participation with political participation and proposed the concept of digital political participation (De Marco et al., 2014).The widespread use of the Internet has changed the governance system of modern countries, with the potential to influence governance systems (Luna-Reyes, 2017).Digital political participation has an impact on network actions.Research has revealed that digital media activities are generously active and associated with the degree of involvement (Lee & Chan, 2016).They play an active role in mobilization, identity building, and organizational transformation (Mercea, 2012).In addition, scholars have examined the positive role of digital participation in promoting the sustainable development of society and creating ''smart sustainable cities'' (Bouzguenda et al., 2019).However, scholars have focused on the participation behavior of a certain group or political participation and digital inequality.Few studies have investigated how to improve digital participation and whether interpersonal communication and online frequency affect digital participation.Accordingly, this study aimed to bridge this gap and explore the main factors affecting digital participation and the possible mediating roles of interpersonal communication and online time.
The theoretical model was constructed based on the social cognitive theory and the theory of planned behavior (TPB).Social cognitive theory posits that individual behavior can be influenced by social networks and personal cognition (Bandura, 2001).In the field of information, social situation is the environmental element within which individuals carry out cognitive activities, and it acts together with personal cognition to drive people's information behavior.Among users, their cognitive and psychological characteristics influence their behavior.Among the most representative cognitive elements are self-efficacy and outcome expectations (Chiang & Hsiao, 2015).In online content production, users are more likely to share content if they have high self-efficacy (Venkatesh et al., 2003).Self-efficacy has a positive impact on users' community participation intention and knowledge sharing (Liou et al., 2016).In addition, social cognitive theory can help explain how sociodemographic variables affect the different forms of collective  participation (Hoffmann et al., 2015).The theory of planned behavior suggests that an individual is more likely to actually participate in a behavior when one's intentions and perceived behavioral are controlled (Ajzen, 1991).Accordingly, we consider them in the theoretical model.

Hypotheses Development
Digital Skills and Digital Participation.Dodel and Mesch (2018) defined digital skills as the basic abilities required to operate digital systems and the skills needed to understand and use online content.Furthermore, digital skills include the capability to respond pragmatically and intentionally, and possess the ability to effectively and efficiently locate content on the web (DiMaggio et al., 2004;Hargittai, 2005).Therefore, scholars have explored the role of digital skills in digital participation and classified these skills.A. van Deursen and van Dijk (2011) classified digital skills into four types: operational, information, strategic, and formal.The improvement of digital skills can strengthen comprehensive abilities, such as equipment operation, opinion expression, and network collaboration.People communicate with others in network environments, which allows individuals to access, share, and store information (Adam et al., 2020) and promotes the possibility of digital search and entertainment.User satisfaction theory (U&G) notes that individuals seek media to fulfill certain needs or goals, which includes the possibility of entertainment and social interaction (Wu & Kuang, 2021).Thus, the following hypothesis was proposed: H1. Digital skills are positively associated with digital participation.
Digital Skills, Online Interpersonal Communication and Digital Participation.The main purpose of participation in digital behavior lies in social interaction, and social function plays a vital role in individual development.With the advent of digital media, internal communication depends on internal media (Hwang, 2011).Moreover, digital media permits many people previously unwilling or unable to communicate the ability to communicate with others.Social media proficiency has influenced interactive and interpersonal skills (Moekotte et al., 2015).The mastery of digital skills means that individuals have more adequate social networking skills and can better communicate with others.A study of Spanish college students showed that university students had higher competence in communicating through interactive presentations and video-image (Va´zquez-Cano et al., 2020).Therefore, the following hypothesis was posed: H2a. Digital skills are positively associated with online interpersonal communication.
The more frequently people communicate with others in cyberspace, the closer their relationship is.Burns et al. (2003) noted that effective communication includes five aspects: present awareness, engagement, interest, public opinion, and understanding.Previous studies have shown that social interaction factors have a significant and positive impact on ICT adoption behavior and feedback interaction with other individuals (Zylka et al., 2015).Individuals involved in network communication become netizens with common interests and hobbies.The changes in the relationship between individuals and likeminded people reveal whether they can promote digital participation.Therefore, frequent contact with the Internet enhances digital participation and encourages people to comment and express themselves more.Therefore, we propose the following hypothesis was proposed: H2b. Online interpersonal communication is positively associated with digital participation.
Previous studies indicated that digital skills may have positive effects on online interpersonal communication, and interpersonal communication may positively affect digital participation.International communication compliance skills have been reported to mediated behavior and offered a practical direction for improving preservice teachers' digital citizenship (Xu et al., 2019).Moreover, qualified digital citizens require good digital skills and digital participation ability; therefore, online interpersonal communication may play an intermediary role in the influence of digital skills on digital participation.As such, this study proposed the following hypothesis: H2. Online interpersonal communication mediates the effect of digital skills on digital participation.

Digital Skills, Online Immersion and Digital Participation.
Online immersion is generally defined as the depth of personal digital use and the time devoted to it (Owston, 2009).Jenkins (2006) argued that online immersion is related to frequent use of digital media and noted the relationship between online immersion and digital skills.Online immersion not only affects the changes of individual physical factors but also emotional factors, which is closely related to the use of the Internet.A study on Chilean teenagers' online immersion demonstrated that digital navigation skills are closely related to social skills and immersion, and attribute toward the use of digital technologies was the most important factor predicting digital immersion, followed by motivation (Leo´n et al., 2022).In the experience literature, participants enter into an extreme version of immersion, losing self-consciousness and experiencing a modified sense of time (Csikszentmihalyi, 1990).This feeling of separation from the real world caused by the immersion gives an inspiration that the online immersion can be reflected by the changes in one's state and the relationship with others.As such, the following hypothesis was posed: H3a. Digital skills are positively associated with online immersion.
A study on public participation in urban planning revealed that participation platforms lacking immersion often lacked practical utility, and immersive planning was a conceptual model to conceive the process of public participation that focuses on the depth and breadth of user experience (Gordon et al., 2011).In addition, a study on online consumers' shopping willingness demonstrated interaction satisfaction and creation value for brands, with a higher sense of experience encouraging consumers to actively share their user experience and improve online interaction (Hamilton et al., 2016).For example, losing one's awareness of time and being deeply involved can occur in online activities such as playing games or surfing information online.As such, this study proposed the following hypothesis: H3b. Online immersion is positively associated with digital participation.
Existing research does not explain whether online immersion is an intermediary factor affecting digital skills and participation.However, according to a common definition of online immersion and digital use, immersion caused by often and extensively using digital devices enhances the possibility of individuals interacting with others in the online environment.It is worth noting that online immersion is more about a subjective psychological feeling than the meatal health repercussions associated with over-use of smart devices (Harwood et al., 2014), which is inconsistent with our research to further enhance citizen digital participation.Therefore, we boldly speculate whether online immersion can be an intermediary factor connecting digital skills and digital participation, precisely because the increasingly optimized network environment makes skilled participants more willing to engage in digital participation.Accordingly, the following hypothesis was proposed: H3. Online immersion mediates the effect of digital skills on digital participation.
In conclusion, this study constructed a theoretical model in which digital skills influence digital participation, treating digital skills (DS), digital participation (DP), online interpersonal communication (OIC), online immersion (OI) as latent variables.Among them, online interpersonal communication and online immersion act as mediating variables of the relationship between digital skills and digital participation (see Figure 2).In addition, we also included demographic characteristics as regulatory variables in theoretical models.Some studies related to the latent variables are listed in Table 1.

Sample and Data Collection Procedure
This study used data from the 2017 Chinese General Social Survey (CGSS), which is the earliest academic survey project in China that has national, comprehensive, and continuous features (Yang et al., 2021).The CGSS used strict scientific sampling to conduct a questionnaire survey on more than 10,000 families in China.The CGSS used face-to-face interview, and the average interview time was about one and a half hours.The CGSS had a set of strict quality control procedures, covering pre-fieldwork, in-fieldwork, and after-fieldwork states.Before the survey began, the participants were informed regarding the survey and its goals.Responses were anonymous, which met the basic survey requirements and ensured content validity and authenticity.During the data input and coding stage, the data must be double input and double coding, and there were several supervisors to check the process.The CGSS comprehensively collected data at the levels of society, community, family, and individual, and is currently recognized as having amassed representative data with scientific research value for academia (Liang & Wang, 2014;Wang et al., 2020).However, it should be noted that the questionnaire content of the CGSS was not consistent between different years.This study mainly examined the survey content related to digital skills and digital participation, so the CGSS2017 data were selected for the study.In CGSS2017, the Part C questionnaire was closely related to the ''network society,'' which fits the topic of this study well.More information about the CGSS and data are available through the China General Social Survey website (China General Social Survey, 2023).
Since most of the data used in this paper were subjective data, men's subjectivity was the factor most strongly affecting the authenticity of the subjective data.For samples with one or more missing values, the authenticity of other data in the sample was also questionable.Hence, this paper adopted the table column deletion method to exclude missing values.That is, in a sample data, whenever there was a missing value, the sample data were deleted.This study collected a final sample of the responses of 1,204 participants to Part C of the questionnaire.

Measurement Instrument
After synthesizing the related items of CGSS questionnaire, we selected four latent variables: digital skills, online interpersonal communication, online immersion, and digital participation.We calculated the model with a total of 16 observation indicators.
Digital skills used four items from A. van Deursen and van Dijk's (2011) definition of digital skills, measured on a 5-point Likert scale.''I could use mobile devices to download and install APP''; ''I could find the information I want online'', ''I will verify the network message''; ''I could express my own opinions online'', anchored with 1 = not at all; 5 = a lot.
Online interpersonal communication used four items adapted from Hudson et al. (2019) scale, measured on a 5point Likert scale 1 = a lot; 5 = not at all, the original data were inversely encoded.The items were: ''I communicate with people who share common interests online''; ''I communicate with my family online''; ''I communicate with my friends online''; ''I communicate with my colleagues online''.
Online immersion consisted of four items adapted from Jennett et al. (2008) scale and from the side of the state changes in daily life, measured on a 5-point Likert scale.''My daily life is being affected''; ''My work is being affected''; ''I have less outings''; ''My family will complain about me'', anchored with 1 = a lot; 5 = not at all, the original data were inversely encoded.
Digital participation consisted of four items ''I participate in online social networking activities''; ''I participate in the online entertainment activities''; ''I went online to get the information''; ''I participate in online transactions'' 1 = not at all; 5 = a lot.

Evaluated Method
This study used a partial least square structural equation model (PLS-SEM) using AMOS 24.0 software.
Structural equation modeling (SEM) is a statistical method to analyze the relationship between variables based on the covariance matrix of variables.It is an important tool for multivariate data analysis.In addition, SEM has many advantages that regression models do not have.SEM can handle multiple independent variables and dependent variables at the same time, and meet the increasingly complex needs of theoretical models in social science research (Hair et al., 2020).SEM can analyze both dominant variables and latent variables (Hair et al., 2019), which include the generally implicit variables in social science research.Moreover, SEM can be applied to small data samples and non-normally distributed data, and its stability has been verified (Reinartz et al., 2009).These advantages make SEM an important statistical method in social science research.

Respondents Characteristics
Table 2 shows the demographics of the sample.Men accounted for 51.9%.All participants were adults, and the age distribution was relatively average.In terms of education level, 62.2% of the participants received high school education or above and had some technology processing ability.Income was classified according to the quintile income published by the National Bureau of Statistics.The results showed that the participants were Note.For personal income, 224 samples of non-income groups, including students, and participants who were unwilling to disclose their income were removed from the demographic statistics.
evenly distributed among middle-income, middle-highincome, and high-income groups; however, the coverage of middle-low-income groups was not wide enough.The above values demonstrated the universality and representativeness of the questionnaire.In addition, 67.4% of respondents spent more than 1 hr online every day, indicating that most respondents had an independent digital use environment.

Data Analysis
Tests of the Measurement Model.Table 3 shows the reliability and validity test results of the measurement model.First, the factor loads of each index in the measurement model should be greater than 0.5 (Hair et al., 2006), and all the factor loads in this study met this condition.Second, the measurement model should test the internal consistency, convergence validity, and discrimination validity of all measurement items.This study used reliability evaluate the internal consistency of the measurement model.The results showed that the Cronbach's alpha (in the range of .734-.856) and concentration ratio (CR; in the range of .726-.861) of the five facets were higher than the recommended value of .6, which demonstrated that the internal consistency of the model was good.Furthermore, this study tested the validity of the model.The results showed that the average variance extracted (AVE) ranged from 0.409 to 0.613 in each configuration, which accorded with the acceptable range of 0.36 (Fornell & Larcker, 1981).Therefore, the convergence validity of this study was supported.In addition, discrimination validity was measured by comparing AVE with the corresponding square correlation estimation of internal structure.As shown in Table 4, all AVE squares for each variable were greater than the correlation of that variable with the other variables.In sum, the model fit, reliability, and validity of this study were supported.
Tests on the Structural Model.The overall fitting degree of the model was tested (Hu & Bentler, 1999) using the chi-square statistics/degree of freedom, root mean square error of approximation (RMSEA), goodness of fit index (GFI), adjusted goodness of fit index (AGFI), comparative fit index (CFI), incremental fit index (IFI), and Tucker-Lewis Index (TLI).The results are presented in Table 5. Due to the relatively large number of samples in this study, the chi-square statistic/degree of freedom (4.492) met the basic requirement, which state that the results for larger samples should be less than 5 (Lefcheck, 2016).Other fit indices met the established requirements, indicating that the model had good fit.
Figure 3 shows the relationship between the structural models and marks the normalized path coefficients between the latent variables.The results showed that digital skills had a significant positive impact on digital participation (b = .734,p \ .001),supporting H1.Testing for the Moderated Effect of Individual Differences.Gender differences among individuals may affect digital skills and participation (Venkatesh et al., 2003).Although the gender gap in home Internet access had largely disappeared in recent years, the digital gap between labor force experiences continues.Men may spend less online time than women due to the pressure of work economy, which may have an impact on their willingness to participate in numbers due to gender differences.Furthermore, more educated individuals more .0.9 .0.9 .0.9 .0.9 .0.9 Note.RMSEA = root mean square error of approximation; GFI = goodness of fit index; AGFI = adjusted goodness of fit index; CFI = comparative fit index; IFI = incremental fit index; TLI = Tucker-Lewis Index.often connect through broadband and spend more time online (Buente & Robbin, 2008).People who have more social resources and higher socio-economic positions are more likely to use the Internet (A.J. A. M. van Deursen & van Dijk, 2015).Thus, differences in individual gender, education level, income, and urban and rural areas impact digital participation.Therefore, this study included some demographic variables in the model as regulatory variables to complete the test of the main effect, grouped different levels of demographic variables as group criteria, and compared the results of regression analysis at different levels under the same variable (Table 7).The results showed that educational level, and urban-rural differences played significant roles in moderating the influencing path of digital skills on digital participation.The amount of personal economic income did not play a significant role in this moderation.

Relationship Between Digital Skills and Participation
Digital skills, as the most important factor in participation in network interactions, are conducive to participation in frequent digital behavior (A.J. A. M. van Deursen & van Dijk, 2015).Researchers believe that the proficiency and use of social media software have a positive impact on information intention and behavior (Bock & Kim, 2002), and the mastery of digital skills leads people to participate in more online behaviors.Such identification ability is conducive to enhancing mutual trust in cyberspace, has a positive impact on knowledge sharing in digital media, and promotes digital participation behavior.Active and highfrequency digital behavior inevitably leads to an increase in network usage time and willingness to participate in active behaviors, such as forwarding, commenting, and publishing, in the process of browsing information, games, and entertainment.When testing for the moderate effects of individual differences, men were found to be more engaged online than women.In cyberspace, men are more willing to participate in collective activities than women, while women tend to participate in non-collective activities, such as healthcare (Caparas & Agrawal, 2016;Son & Lin, 2008).Higher educational levels were associated with greater digital participation, which indicated that the development of network societies cannot be separated from the improvement of knowledge and learning level.Urban  residents were more exposed to networks than rural residents and were more willing to participate digitally.Furthermore, cities had more social resources and interpersonal communication than rural areas, and the Internet was used more frequently there.Finally, this study found that the regulatory effect of personal income differences was not significant, which indicated that the access network for digital participation no longer required high economic costs, and different income groups could have effective digital participation.

Relationships Between Digital Skills, Online Interpersonal Communication, and Digital Participation
Previous studies have shown that people' attitudes toward digital use change their willingness to use the Internet (Fishbein & Ajzen, 1975) and are influenced by relatives and friends (Huang, 2013).However, few studies have investigated how online communication and changes in personal life jointly affect digital participation.H2a and H2b examined the influence of digital skills on online interpersonal communication, and of online interpersonal communication on digital participation, respectively.The model shows that there was a positive relationship between digital skills and interpersonal communication, and closer communication (whether online or offline) had a positive impact on digital participation.This indicated that communication between people in the network promoted intimate relationships with others and participation and expanded the scope of communication.In addition, online interpersonal communication played an intermediary role in the influence of digital skills on digital participation, as argued by H2.Individuals with higher digital skills had a greater possibility of digital participation through online communication with others, which explained the mechanism whereby, interpersonal communication to enhance participation.

Relationships Between Digital Skills, Online Immersion, and Digital Participation
The results demonstrated that digital skills had a significant influence on online immersion, and online immersion influenced digital participation.Mastering digital skills made individuals feel immersed in cyberspace, which impacted their life and work, as suggested by H3a.This was in line with Leo´n et al. (2022) and confirmed that individuals with higher digital skills were more willing to spend time in cyberspace.Moreover, as proposed by H3b, increased immersion enhanced the possibility of digital participation.Due to the openness of cyberspace, even the smallest hobbies could get attention and communication on the digital platform.Among digital skills, web browsing, business transactions, and software downloading led to the deepening of communication with others and enhanced intimacy in this relationship.
The impact of immersion expressed in this study was not entirely positive, and it is still necessary to guard against the negative impact of online immersion on individual health and communication relationship.Due to the existence of Internet addiction, groups susceptible to immersion, including teenagers, may neglect their attention to real communication.In addition, as argued by H3, online immersion played an intermediary role in the influence of digital skills on digital participation.Therefore, in addition to online interpersonal communication, individuals with high-level digital skills could enhance the possibility of digital participation through immersion in the Internet.

Theoretical Implications
This study was based on the data from China's national survey and aimed to examine digital participation.
Unlike previous studies on the numerical differences of individuals divided by age, there are significant differences in regional economic development, population composition, and many other aspects in China's environment.Therefore, this study built a model of the influencing factors of digital participation from a holistic perspective.
Furthermore, this study verified the influence of digital skills on digital participation presented in previous studies (De Marco et al., 2014).This was the first study aiming to explain the influence of individuals' communication with others in cyberspace and immersion on digital participation.Moreover, this study demonstrated that online communication and immersion played an intermediary role in the influence of digital skills on digital participation, which provided a preliminary blueprint for further research.
In addition, the study revealed that demographic disparities, including gender, educational attainment, and rural-urban disparities, influenced digital participation, which was consistent with previous studies on the digital divide (Buente & Robbin, 2008;A. van Deursen & van Dijk, 2011).However, the difference in personal income is not a significant factor, which shows that the Internet construction in China has been fully popularized, the digital skills of citizens are close to maturity, and the opportunity and frequency of digital participation are continually increasing.Therefore, this finding transcends the conclusion that originally income affects the impact of digital skills on digital participation.It further shows that the future improvement of citizens' digital participation should start from the improvement of participation experience and participation literacy, rather than device access.Overall, this study makes theoretical contributions to the factors influencing digital participation and important role of digital skills.Moreover, it expands our understanding of how to better promote digital participation and the key role played by individual realistic factors.

Practical Implications
The results of this study have implications for policy related to digital participation.For government departments, it is particularly important to make good use of network media as a carrier to enhance active participation.
Digital skills are the first condition for active digital participation.In the process of encouraging digital participation, the government should first cultivate digital skills.From a guiding point of view, we should strengthen the popularization and guidance of Internet use in remote areas and among middle-aged and elderly people, allowing more groups can to be included in digital participation.
Furthermore, the research results demonstrated that communication with others through online media can affect the achievement of digital participation, and digital skills can make communication perception more intimate closer.Therefore, it is important for policymakers to pay attention to the good and orderly interactive content of people in cyberspace and build social platforms that facilitate communication.
Moreover, digital platform creators should pay attention to the positive impact of online immersion on digital participation, which means that a carrier that can help participants immerse themselves online for a long time is crucial for active participation.Platform builders can improve users' online use time by creating page settings that meet the esthetic needs of target groups, simple and quick operation interfaces, and detailed information arrangements, thereby facilitating active digital participation.In addition, excessive immersion can lead to online addiction.Therefore, the government should build a scientific and healthy Internet interaction platform to enrich the forms of Internet interaction between individuals and others.
In addition, considering the intermediary role of online interpersonal communication and immersion, they can no longer be considered isolated factors affecting digital participation, and digital participation can no longer only be considered as being related to digital skills.Therefore, it is necessary to establish a good network social organization to form ideal communication and interaction systems.The willingness of Chinese online communities to participate continuously depends on satisfaction and affective commitment (Jin et al., 2010), which shows that public participation is more influenced by individuals than by organizational mobilization.This suggests that the government should not only continue to cultivate social organizations but also urge more individuals to participate digitally by means of interpersonal communication.
Finally, the government should strengthen the construction of government information disclosure, allowing individuals to conveniently obtain the required participation information, improve the asymmetric status of information, and facilitate effective participation.At present, the Chinese government has begun to implement the relevant measures of digital government, and governments at all levels have launched digital participation platforms.However, the connection mechanism between public participation and public affairs decision-making is not perfect, which leads some people to believe that participation cannot change the status quo.This requires the government to reshape the organizational boundary, reform and innovate the allocation mode of public resources, and improve the transparency of government affairs operation.

Limitations and Future Research
This study had some limitations.First, the changes in China's Internet development and the promotion of digital government construction in recent years mean that data from 2017 lag behind.At the same time, since CGSS did not conduct a survey on the content of the ''network society'' every year, the study was temporarily unable to compare and analyze the data between different years, and this part would be supplemented after the future data were published.Second, this study was mainly aimed at adults, as data on the digital participation of adolescents are insufficient.Future research should focus on the digital participation of adolescents, who have the most access to the Internet.Third, this study did not investigate a specific group, and future research could be targeted from the economic, education, and other aspects of classification research.Finally, research on the negative impact of digital participation is insufficient.Facing the emergence of the digital divide, future research should include nontraditional factors, such as discourse expression and subcultural communication.

Conclusion
This study examined the factors that enhance digital participation in many respects, revealing the key role of digital skills in improving digital participation and proposed strategies to strengthen digital skills training and education and optimize digital platform page operation.In addition, this study examined the positive effects of online interpersonal communication and online immersion on digital participation and demonstrated that online interpersonal communication and immersion played an intermediary role in the process from digital skills to digital participation, which has enriched our understanding of the changes of communication relationship and physical and mental perception produced by individuals through networks.Finally, this study indicated that gender, education, and regional differences influenced the results; therefore, it is necessary to further improve the measures of digital platforms in content guidance and target group selection.

Figure 1 .
Figure 1.The overview of the study.
Note.DS = digital skills; OIC = online interpersonal communication; OI = online immersion; DP = digital participation.Digital skills for online personal communication (b = .182,p \ .001)and online immersion (b = .195,p \ .001)had significant positive effects, supporting H2a and H3a.In addition, online personal communication (b = .087,p \ .01)and online immersion (b = .109,p \ .001)had significant positive effects on digital participation, supporting H2B and H3B.Results of Mediation Effect.According to the Bootstrap method recommended by Hair et al. the study estimated 2000 bootstrap samples in 95% CI.Table 6 displays the test results in bias-corrected and percentile modes.The indirect effect of online interpersonal communication represents the path of digital skills !online interpersonal communication !digital participation.The indirect effect of online immersion represents the path of digital skill !online immersion !digital participation.The results showed that online interpersonal communication mediated the relationship between digital skills and participation (indirect effect = 0.022; 95% CI [0.012, 0.039], [0.011, 0.036]).Online immersion mediated the relationship between digital skills and digital participation (indirect effect = 0.018; 95% CI [0.006, 0.035], [0.006, 0.032]).Moreover, the value range of the direct effect of the model in the 95% CI ([0.649, 0.863], [0.648, 0.862]) did not include the value 0. Therefore, online interpersonal communication and immersion had significant mediating effects on digital participation and were partial mediators, supporting H2 and H3.
ICT = Information and Communications Technology; TPB = Theory of planned behavior; CGSS = Chinese General Social Survey; PLS-SEM = partial least squares structural equation model; SEM = structural equation model; CR = concentration ratio; AVE = average variance extracted; RMSEA = root mean square error of approximation; GFI = goodness of fit index; AGFI = adjusted goodness of fit index; CFI = comparative fit index; IFI = incremental fit index; TLI = Tucker-Lewis index

Table 1 .
Description of Some Studies Related to the Research.

Table 3 .
Convergent Validity and Internal Consistency Assessment of Latent Variables.

Table 5 .
Measurement Model Fitness Test Results.

Table 6 .
Results for Mediating Model.

Table 7 .
Results for Moderating Effect.