Gaming social capital and civic involvement among Russian MMORPG players

Ðàçìåùåíî íà http://www.allbest.ru/

FEDERAL STATE AUTONOMOUS EDUCATIONAL INSTITUTION

FOR HIGHER PROFESSIONAL EDUCATION

NATIONAL RESEARCH UNIVERSITY HIGHER SCHOOL OF ECONOMICS

Faculty St. Petersburg School of Social Sciences and Area Studies

Bachelor's project

Gaming social capital and civic involvement among russian mmorpg players

Daria Stanislavovna Mikhalchuk

Saint Petersburg 2020



Table of contents

Introduction

1. Theoretical background

2. Civic Engagement and Its Classical Predictors

3. Literature review

4. Methodology

5. Data preparation and description

6. Results

Discussion

Limitations

Future work

Conclusion

References

Appendices



Introduction

Online games have been gaining in popularity for years. They have become a full-fledged, extensive part of life for many people, a virtual space where they communicate and interact with each other. Online games are treated as a new environment for individuals to mediate social relationships (Nardi and Harris 2006; Voida and Greenberg 2009). Inside the game, players perform actions typical for everyday real life so that online games can be considered as “models of the real world” (Castronova et al. 2009). Moreover, online games can be examined as “third places” (Oldenburg 1997) because they have become as important as home and work, i.e., the “first” and “second” places (Steinkuehler and Williams 2006).

In addition to gaming, regular players collaborate, create their own digital communities and translate their game life into other forms of activities. In 2016, users of the Reddit.com platform who played the Sid Meier's Civilization V created the community Democraciv for managing one of the states in the game (Habr.Com 2016). Six political parties appeared in the community, the party members were Reddit users, and the community supported the democratic regime. Moreover, there also existed three branches of government, power distribution and even the Constitution.

Online games are highly popular in Russia; in 2018 alone, 65 of 144 million Russian people played online games, which is 45 percent of the population, and the total revenue of the Russian game market accounted for 1.7 billion US dollars, making Russia the 11th biggest games market in the world (Newzoo Russia Games Market 2018). Around the globe, the online gaming industry is gigantic: the number of players has reached 877.3 million in 2020. By 2022 this figure is predicted to rise to 960 million online players (Statista 2020). Such popularity of online games makes it important to study what impact they have on the players' lives.

This research focuses on the Massively Multiplayer Online Role-Playing Games (MMORPGs). An MMORPG is a game environment which can host thousands of people simultaneously in the same virtual online world or space (Barnett and Coulson 2010). The distinctive feature of MMORPGs is that people there interact with the game field and with other players by creating the fantastical avatars or heroes - in-game representations of players which can be either similar or dissimilar to them (Blinka 2008). The most popular MMORPGs in 2019 by the number of players were The World of Warcraft, The Elder Scrolls Online, Guild Wars 2, Final Fantasy XIV, Runescape, etc. (BeStreamer.Com 2019).

MMORPG is an “old”, well-researched genre of online games. MMORPGs have been widely researched from the perspective of social capital, as a social space with numerous interactions and collective play (Ducheneaut and Moore 2004; Shen and Chen 2015; Stokes and Williams 2018). However, there is a gap in studying how the social ties which were created through the gameplay affect the daily, real life of players that is detached from the gaming world. Moreover, there are persistent myths about online gaming as a pervasively negative, harmful experience and a stimulant of aggressive behavior. Anderson and Bushman (2001), Ng and Wiemer-Hastings (2005), Hollingdale and Greitemeyer (2014) in their works argued that violence in video games promotes aggressiveness, negative feelings and attitudes. With my research, I want to put these results to test on empirical evidence.

There is a debate on the time displacement effect of games. On the one hand, online gaming distracts people from everyday offline actions as it takes a large amount of time to play video games (Williams 2006; Shen and Williams 2011; Kowert et al. 2014). On the other hand, such virtual space is an opportunity to accumulate social capital and gain social ties which may bring some valuable resources to players (Zhong 2011; Molyneux et al. 2015).

Thus, I want to contribute to the field by exploring the potentially positive impact of multiplayer online games on accumulating in-game social capital as well as on the civic involvement of players.

The paper focuses on Russian MMORPG players. My research question is the following: Can gaming social capital, i.e., social capital obtained in online gameplay, be positively related to players' involvement in the civic life of their community or the country? The aim of this paper is to investigate the relationships between the social capital accumulated through gameplay and the engagement in civic actions which are not connected to online games or which are performed outside the virtual gaming world, within the context of Russian MMORPG playing community. In this research, I want to examine and test the theoretical path from playing multiplayer online games to gaining gaming social capital and then to civic engagement. In this paper, I use the terms “civic involvement” and “civic engagement” interchangeably.

The paper accomplishes the following empirical research goals:

to reveal whether multiplayer gaming experience is positively related to gaming social capital;

to investigate the association between gaming social capital and the level of civic engagement;

to examine whether the accumulation of the gaming social capital can transform into non-civic involvement.



1. Theoretical background

Gaming Social Capital

This research is grounded in social capital theory. Studies of social capital have mostly followed one of the two research traditions, focusing either on the individual-level social capital (Coleman 1988; Granovetter 1983) or on the collective social capital (Putnam 2000). According to Coleman (1988), social capital is a set of resources which are received through social links between people. In other words, social capital describes information, knowledge, obligations, and norms available to the person from his or her relations (Coleman 1988). Lin (2001) defines social capital as the resources that are accessed in rational action that give better outcomes for individuals.

At the individual level, social capital could be operationalized as resources embedded in individuals' social networks. At the collective level, social capital creates benefits for the community such as trust, cooperation, reciprocity and collective action (Coleman 1988; Putnam 2000). By using weak ties (Granovetter 1983) and “friends of friends” (Boissevain 1974), members of the community can gain privileged access to information, skills and opportunities.

Social capital is a source that weaves individuals in a community based on shared interests, ideas and goods (Etzioni 1996). Social capital rests on reciprocal relations (Coleman 1988; Paxton 1999) and social trust (Putnam 1995). The indicators of social capital include social support and common information resources (Mathwick et al. 2008).

This paper focuses on gaming social capital, a concept that narrows down the general idea of social capital discussed above to the online gaming space. I consider multiplayer online games as a social environment for connecting individuals, their shared activity and interactions. Moreover, this research takes into account another aspect of gaming social capital - a sense of belonging to a community which can be defined as a degree of social connectedness with the social group that gives an opportunity for people to work together and cooperate (Molyneux et al. 2015).

Time-Displacement Theory

Online games are often presented as a distracting environment that may devour time, thus subduing any offline social activities. This is a problem of displacement which means that spending time on some new activities can displace the time which was usually dedicated to other activities (Nie 2001). Displacement occurs in different contexts. For instance, online gaming can displace time usually spent on interacting and meeting with family and friends, or face-to-face communication in general (Cummings and Vandewater 2007) as well as on participating in the life of community or the country.

There is a stereotype that gamers isolate themselves from the real world and do not communicate with people (Williams, Yee, and Caplan 2008). However, most players do not abuse gaming time. Moreover, online gaming can help to create new social ties (Cole and Griffins 2007). Although the displacement effect of online games can be observed, it does not necessarily apply to MMORPGs where teamplay is promoted.

The opportunity to join or create guilds provides MMORPG players with the sense of belonging to a community that may help to gain gaming social capital (Zhong 2011). This happens because players work together in the guild, accomplish the same goals, so they share their online space with others that enhances their psychological involvement (Biocca et al. 2001). However, the problem of displacement is still a debatable question that needs further investigation. While there is a research-based claim excessive online gaming can lead to higher chances of depression and poorer offline relations (Liu and Peng 2009), the results of other papers show that the MMORPG gameplay can help nourish stronger and broader social ties (Trepte, Reinecke, and Juechems 2012).

The social component of MMORPGs as well as players' collective behaviors and their motivations can be a decisive moderator between time spent playing and social relationships (Zhong 2011). The more an individual plays a cooperative game in a guild, the more gaming social capital will be accumulated. Alternatively, the more an individual plays by himself or herself without entering the collective gameplay, the more time on social interactions will be replaced by time spent on gameplay; such players have poorer social capital (Zhong 2011; Castillo 2019; Reer and Krämer 2019).

2. Civic Engagement and Its Classical Predictors

Civic engagement is widely used as an indicator of political action. It is measured by evaluating the individual's knowledge of political events and state activities as well as by counting the frequency of his or her political actions (Verba, Schlozman, and Brady 1995; Smith, Schlozman, Verba, and Brady 2009). Several studies distinguish between political and civic engagement. Political engagement aims at making an impact on the authority's actions and on the draft legislation (Verba, Schlozman, and Brady 1995), whereas civic engagement is grounded in resolving and dealing with the problems at the community level (Park and Perry 2008). It means that civic engagement consists not only of voting and volunteering actions, but also includes participation in meetings and rallies. Despite the fact that civic actions tend to move online as social media gain in popularity and functionality, civic involvement is still measured by the type and number of actions rather than attitudes, opinions or considerations that tend to go along with particular actions (Gordon and Baldwin-Philippi 2014). Some researchers even criticized online civic action because it allegedly lacks the impact on community life when compared with offline voting, donating money to charity, etc. (Shulman 2009).

In empirical studies of civic engagement, various indicators are used. Son and Lin (2008) researched the relationships between social capital and civic engagement where they defined civic engagement through actions such as volunteering, attending political meetings, signing a petition, and others. Keeter et al. (2002) proposed nineteen core indicators of engagement which are divided in three categories:

(1) civic indicators (volunteering, active membership in a group or organization, participation in fundraising events, participation in charities, etc.);

(2) electoral indicators (voting, displaying signs or stickers, volunteering for political organizations, etc.);

(3) indicators of political voice (contacting officials and/or media, protesting, signing or creating petitions, etc.).

All these indicators have traditionally related to offline actions but they can be extended to the online social life as, for example, signing petitions nowadays is also an online activity.

3. Literature review

The MMORPG Environment

Broadband Internet connection has greatly enhanced the gaming process by allowing players to play online games together, in the same virtual space either in pairs and small groups or even in hordes of hundreds of people. MMORPGs are remarkable for engaging players in social interactions as many activities or tasks in this type of game should be performed in a guild in order to increase the efficiency of some actions, to gain achievements or sometimes even to start or finish the playing session (Ducheneaut et al. 2006). A guild is a self-organized or randomly organized group of players in a massively multiplayer online game that helps them to perform coordinated actions, to pass missions and to solve challenges (Chen, Sun, and Hsieh 2008). Not all in-game actions demand cooperation. The guild is needed to achieve some short and long-term goals within the game, for example, to kill monsters, to explore the parts of the gaming world, to share items, to solve quests and to participate in the battles (Nardi and Harris 2006). Cooperative playing in guilds provides players with more social benefits as compared to individual gameplay or to the competitiveness within the game (Velez et al. 2016).

MMORPGs are particularly renowned for harnessing social capital within the community through incentives for social interaction. They provide individuals with the full alternative virtual world where players should improve their skills, solve problems and challenges and learn from each other to achieve some goals. MMORPGs not only stimulate this sort of social interactions, but also manage players to organize guilds and show the leading role or join existing groups, to cooperate, and to make contributions for their communities, representing themselves through the individually created avatars (Steinkuehler and Williams 2006; Zhong 2011). Such in-game actions can be considered as endorsing active community participation and developing the feeling of belonging to the group and trust between members (Ratan et al. 2010).

When people learn how to collaborate and work together in the virtual game world, they can apply this experience and these skills in real life and contribute to the activities of various civic organizations. The game environment consists not only of the in-game activities between players, such as communicating with each other by voice or text messaging, creating guilds, etc. (Manninen 2003), but also includes social actions on the forums or fan pages dedicated to different MMORPGs. For example, an MMORPG The World of Warcraft has a page on the social networking sites such as Reddit (https://www.reddit.com/r/wow/) and Vkontakte (https://vk.com/worldofwarcraft) where people can communicate, post news, and share their knowledge, hints and opinions in comments. There is also the Russia-speaking Steam Community (https://steamcommunity.com/groups/ru_steam) which publishes and discusses various news related to Steam, Valve's A game-developing and distributing company based in the USA. distribution service, in Russia.

Impact of Online Games on Social Capital

MMORPGs have a mixed reputation when it comes to the social capital of players. Some researchers state that excessive online gaming leads to losses in social life as it takes away from social relations and reduces the time people spend on the offline activities (Williams 2006; Shen and Williams 2011; Kowert et al. 2014). Moreover, excessive MMORPG or violent games playing, under some circumstances, can lead to aggressive behavior, depression and poor psychological well-being (Guinn, Bickham, and Rich 2011; Kirby, Jones, and Copello 2014; Turel 2020). In addition, cyberbullying attacks have been reported to occur in the MMORPGs (McInroy and Mishna 2017).

However, there is plenty of other research that shows how online games can be a joint activity for the people and can help to extend the existing network of friends and create new ties (Cole and Griffiths 2007; De Grove 2014; Domahidi, Festl, and Quandt 2014). What is more, Cole and Griffiths (2007) state that gaming activity may be a good “soil” for building strong emotional and life-long friendships. Additionally, some authors argue that online gaming can motivate the growth of social capital (Steinkuehler and Williams 2006; De Grove 2014; Eklund and Roman 2017). There is no sweeping conclusion in the literature on this issue.

Gaming Social Capital and Civic Engagement

Researchers from various fields of study have examined and tested the existence of the spillover effect of games when the behavior and actions that were formed in the online space may transform to the offline reality (Williams 2010; Xanthopoulou and Papagiannidis 2012). Several studies argue that social links and interactions in collaborative online gameplay may have a positive effect on the creation and development of the community as well as on the social behavior of players. An online game can be a tool for molding and stimulating civic involvement as online gaming brings up some community formation aspects and teamwork (Gordon and Baldwin-Philippi 2014). In this vein, some researchers claim that immersion in multiplayer games can increase the quality and the quantity of players' civic engagement (Zhong 2011; Dalisay et al. 2014; Molyneux et al. 2015; Stokes and Williams 2018). Zhong (2011) has found that the gaming social capital obtained through collaborative gameplay has a positive impact on civic engagement; the same association was observed with the youth who were more likely to engage in the civic life when they had more gaming experience and spent more time on the sites related to gaming (Kahne, Middaugh, and Evans 2009).

Hypotheses

Taking all this into account, I want to investigate the relationships between social capital obtained through gameplay in MMORPGs and civic engagement in the following hypotheses:

H1: Multiplayer gaming is positively associated with gaming social capital of players among the experienced players.

H2: High gaming social capital is positively associated with civic engagement outside the game environment.

H3: High gaming social capital is positively associated with non-civic engagement outside the game environment.

H4: Excessive MMORPG playing displaces the time for offline social activity, i.e. at excessive rates of online playing, gaming social capital is negatively related to civic engagement.

4. Methodology

Operationalization

Multiplayer gameplay is defined in the context of online gaming as interactions between players in the same virtual environment where players achieve collective goals together, collaborate on solving problems within the game and learn tactics from each other (Ducheneaut and Moore 2004).

Gaming social capital means social ties among gamers and interactions which arise during multiplayer gameplay (Molyneux et al. 2015).

Civic engagement refers to the citizens' involvement in public affairs, participation in the life of their community, the intention to resolve community problems; it consists of political and non-political activities (Park and Perry 2008).

Excessive playing can be operationalized as the overuse of online games (more than twenty hours per week) that may lead to negative and harmful consequences for players, sometimes referred to as “video game addiction” (Griffiths and Davies 2005).

Data collection

For this research, I have conducted an online survey which was conducted with Russian MMORPG players. The language of the survey was also Russian. With the help of the online survey, I have accessed different gaming communities and people with different levels of civic engagement and gaming social capital. Moreover, the online survey gave me a possibility to collect free comments from my respondents, collect their feedback questions and reflections on the survey (some comments from the respondents are provided in the Appendix section).

As a measure to prevent a high non-response rate, I developed the following reward system for respondents. The introduction to the survey said that I would randomly choose ten respondents who would receive games-related gifts from me. I have not conducted the raffle yet, but I will organize the prize competition in the mid of June as 40 percent of respondents left their contacts for taking part in the lottery.

Moreover, 23 percent of respondents shared contacts (usernames in Discord, social network accounts, or emails) of their friends who likewise played MMORPGs and who could potentially be my respondents. For such an option, when I wrote the potential respondents directly, the response rate was about 45 percent.

My target population are Russian players of MMORPGs who have at least 100 hours of gaming experience and who have three or more years of experience playing MMORPGs. Since civic engagement is measured over the last three years, I am not interested in respondents who quit online gaming some time ago or started playing after the registered civic activity. I have also included these requirements in the description of the survey.

For the data collection, I used the free Google Forms tool, where I have seven sections for the respondents with 28 closed questions in total. Also, I have two optional open questions: the first one is about giving the contacts which I will use in case the respondent wins in the lottery, the second question is the request of their friends' contacts who could also take part in my survey. To protect the respondents, all the data are anonymized. Their personal data such as emails will not be stored.

Before conducting a survey, I pre-tested two versions of the questionnaire: in the first one I showed the answers as a scale that is a built-in function of the Google Forms service, while in the second one the answers were written as a text for each point. I tested both versions on ten respondents who said that the answers with scales are better perceived. When all the questions for my measures were formulated, I ran a pilot survey on twenty respondents and collected their feedback. Then I made some corrections and started to post my survey in thematic communities on VK (https://www.vk.com/) and Reddit (https://www.reddit.com/). I have chosen these online communities because, first, they are the most numerous ones for the Russian-speaking MMORPG community, second, they refer to official games' pages and, third, they are free and open for followers or users. On Reddit, users can post openly, without paying. By contrast, in VK, the majority of communities are run by administrators and I had no opportunity to post something by myself. I used three options when I wanted the community to post my survey: the first one was to message the PR-managers or page administrators individually and ask them for either a paid advertising post or a free post in exchange for sharing part of my research results that would be interesting for players. The second option was to message the community account; the third option was to use the “Suggest news” button or just post the survey on the page “wall” if I had such an opportunity. I used the first option more frequently than the other two: some administrators, PR-managers or supervisors of the communities were interested in the research and found it worth publishing, but others, for example in big communities with about 60 000 followers or more, agreed to post my survey only as an advertisement. I paid for such offers from 350 to 2 500 rubles ($5-34 US dollars) per post. All the communities of my choice have approximately 20 000-300 000 participants. As a result, I have received 339 responses to my survey. All answers were collected automatically in the Google Tables.

Measures

For measuring multiplayer gameplay, respondents were asked how often they played MMORPGs in a group or guild online and how many hours their average collective playing sessions lasted.

To measure gaming social capital, I have modified part of the gaming social capital scale provided by Molyneux et al. (2015). The measures which I further combined in an index focused on the sense of belonging to a community as this aspect was discussed in the literature and can be applied in a gameplay environment (Williams 2006; Zhong 2011; Dalisay et al., 2014). I used the following four statements: “I feel close to the people I play MMORPGs with”, “I feel lonely when I do not play MMORPGs for some period of time”, “I have offline relationships with the people I play MMORPGs with”, and “I feel like I am part of a community of players”. The answers to these questions were provided on a 1-7 Likert scale where 1 is “strongly agree” and 7 is “strongly disagree”.

Civic involvement, or civic engagement (the terms are used here interchangeably), is a dependent variable in my study. I measured it based on the definitions and the construct of civic participation by Keeter et al. (2002), with some adjustments. The respondents were asked to answer questions about their participation in political and social activities over the last three years (except for voting which is measured for the previous year only). Below is a list of questions which I used to measure the civic engagement of my respondents:

Did you vote in elections in 2019-2020?

Did you donate money to charity over the last three years?

Did you sign paper petition(s) in the authorities over the last three years?

Did you sign online petition(s) over the last three years?

Did you take part in permitted rallies/protests/demonstrations/processions over the last three years?

And in unauthorized rallies/protests/demonstrations/processions over the last three years?

Were you a volunteer at any event over the last three years?

According to my third hypothesis (H3), I also included questions which measure the non-civic form of engagement, i.e. professional or work engagement, participation in festivals, contests, etc. I measured non-civic engagement with the questions:

Would you like to participate more in solving problems related to environmental issues, fighting domestic violence, unemployment, or drug addiction, or developing arts and culture?

Did you organize professional or leisure events during your studies and/or at work over the last three years?

Did you participate in various contests/festivals/olympiads over the last three years?

In my study, I control for the respondent's sex, age, city of residence, the MMORPGs which a respondent plays, the frequency of MMORPG play (hours per week) and the experience of playing MMORPGs (in years). For all the questions, I provide respondents with the list of possible answers and an opportunity to write their own answer if they want to.

Moreover, following the literature, I control for interest in politics and social news as research has found a positive relationship between interest in politics and civic engagement (McLeod et al. 1996; Shah et al. 2001; Gil de Zúñiga, Copeland, Bimber 2014). I measured interest in politics variable using the following questions:

How interested are you in the news on politics and society (i.e., elections, new legislation, innovations, domestic and foreign policy, social development, current situation in the country/world)?

How often do you discuss political and/or social issues, news with other people?

How often do you share political and/or social news with your friends on social networks?

As the whole world has been affected by the COVID-19 pandemic, I have adapted my survey to the situation of the imposition of lockdown in the Russian Federation. I added three questions to measure what impact the COVID-19 pandemic has had on the gaming community and their political attitudes:

Did you start playing MMORPG more often during the COVID-19 lockdown?

Have you started to communicate ever more closely with players in MMORPGs during the COVID-19 lockdown?

Evaluate how well the Russian government copes with the spread of the COVID-19 virus.

Reflection on data collection

The data collection process lasted for two months (from the end of March to early May, 2020), and my target population was active and fast in filling in the questionnaire. I got the majority of responses in one or two hours after I or the administrators posted the news about the survey on the community page. The reasons for this activity of players may be their interest in the questions and in the process of sharing their experience, but also the current COVID-19 pandemic and the lockdown, under which people must stay at home, thus, having more free time, for example, for surfing the Internet and filling in the questionnaires. Notably, the respondents from VK were much more active than the respondents from Reddit as they left more comments under my post or even re-posting the survey on their personal pages. I think that, since I made the majority of posts in VK through the administrators of the communities, players trusted this survey more than if I had made the post by myself.

The data collection process was accomplished with the comments of the respondents who wrote their reflections, opinions or further questions below the post with my survey. Some comments can be found in the Appendix section (Figures 1-4). Overall, the respondents inquired about the specific purposes of my study, where they can find my research results, and the exact name of my university; others asked to show them the statistics of the game use (Figure 2). I faced some small difficulties with a few respondents who did not trust my survey at all and wrote that I should be a “member of the government and want to collect information about voting and writing petitions in such a “sly” way” (e.g. Figure 4). These respondents not only refused to participate in the survey just ignoring it, but even reached me in private messages to express their opinions and say that I misinformed them. However, when such comments were posted for public view, other respondents engaged in controversy and justified the survey. All in all, the respondents divided into two groups: those who were interested in the research, filled in my questionnaire, left their contacts for participating in the lottery and those who claimed that the government collects statistics for their own aims through such surveys. This happened even though I wrote in the description of the survey for the respondents that the data were collected for the academic research, the questionnaire was anonymous and that by filling in the survey meant that the respondents agreed that their answers would be used for making general conclusions.

5. Data preparation and description

All the data preparation and analysis were done in the RStudio with the help of the R programming language. To prepare the data for analysis, I checked for duplicates and NAs in my dataset. No duplicates were found. There were also no missing data and no noise data. One reason is that all my questions were marked as compulsory ones; the respondents could not skip questions and could not submit the form without answering, but, as I mentioned before, I had two optional questions where respondents could leave their own or their friends' contacts.

After this, I checked the classes of variables and manually cleaned the data to exclude responses coming not from Russia (for example, I have respondents from Armenia, Minsk, Belarus, Latvia and Ukraine) who are not my target audience. Moreover, I excluded 11 respondents who had played MMORPGs for less than three years in total because I measure the civic engagement of respondents over the last three years. By cleaning the entries where the gaming experience was less than three years, I measured the civic engagement only for those players who have had gaming experience before the expected time frame for measuring civic engagement.

I also reversed the scales for some variables, categorized the responses which could be sometimes the own written answers of the respondents (e.g., the city of residence, frequency of playing MMORPGs per week where there was the possibility to fill in “other” category) and translated all responses into the English language. After the data preparation process, I had a dataset with 321 responses that I analyzed.

My typical respondent is a male aged 21-25 years old, located in Moscow, who plays MMORPGs 6-10 hours per week, has 7-8 years of gaming experience and the most-played MMORPG of his choice is The World of Warcraft. The graphical visualizations o some distributions can be found in the Appendix section.

Figure 5 shows that the sample consists of 74 percent of male respondents (n = 237) and 26 percent of female participants (n = 84, only two options were provided). As I measured age with predefined age groups, I have the following right-skewed distribution of age (Figure 6): 36 percent of the sample are aged from 21-25 years old (n = 114), 28 percent are aged from 18-20 years old (n = 91), 20 percent are in the 26-30 age group (n = 64), 7 percent are aged from 31-35 years old (n = 22), 3 percent accounted for participants from 36-40 years old (n = 9), 2 percent were aged from 41-45 years old (n = 8) and I have 2 percent of respondents whose age is less than 18 years old (n = 8) and 2 percent of participants aged more than 45 years old (n = 5). Figure 7 shows that the majority of my respondents are from Moscow (n = 99) and Saint-Petersburg (n = 50), 31 percent and 16 percent, respectively; next most popular cities were Novosibirsk (n = 19), Samara (n = 15), and Yekaterinburg (n = 14), while other 39 percent of respondents named 59 different Russian cities (n = 124). As for the playing hours spent in MMORPGs, Figure 8 shows that the 35 percent of participants play from 6-10 hours per week (n = 111), 24 percent play 11-15 hours per week (n = 76), 16 percent of respondents play excessively more than 20 hours in a week (n = 52), 15 percent spend 1-5 hours per week (n = 47) and the remaining respondents are accounted for by 16-20 hours per week (n = 35). Figure 9 demonstrates that 44 percent have 7 to 8 years of gaming experience (n = 142), 37 percent have played MMORPGs for 5-6 years (n = 119), 8 percent have playing experience from 3-4 years (n = 27), 6 percent played more than 10 years (n = 18) and 5 percent have from 9-10 years of MMORPGs gameplay experience (n = 15). The top most-played MMORPGs (Figure 10) were The World of Warcraft (n = 99), The Elder Scrolls Online (n = 84), Final Fantasy (n = 43), Lost Ark (n = 21) and Blade & Soul (n = 5); in total, respondents named 46 various most-played MMORPGs.

For the regression analysis, I merged some categories for the most-played MMORPGs and the city of residence as I have too many different responses there which cannot be displayed in the model. I divided all 64 cities into three groups based on their popularity: “Moscow”, “Saint-Petersburg”, and “Other”. Following the same procedure, I merged the most-played MMORPGs leaving the two most popular games in my sample, The World of Warcraft and The Elder Scrolls Online, and treating the remaining online games in my sample as a category “Other”.

As for the COVID-19 related questions, 71 percent answered that they started playing online games more often (n = 228) while nothing changed for the rest of the sample, 29 percent. Also, 44 percent of respondents began to communicate ever more closely with other players in MMORPGs (n = 142), while 9 percent of respondents pointed out that they would communicate more if that was possible (n = 30) and nothing has changed for 17 percent of players (n = 56). Last but not least, my respondents evaluate how well the Russian government copes with the spread of COVID-19. The distribution of answers on the question “How well the Russian government copes with the spread of COVID-19?” is represented in Figure 11. The variable is fairly normally distributed with a heavy left tail; the average evaluation is 4.6 (SD = 2.42).

The most popular answers were “6” and “7” (n = 109) on a scale from 0-10.

Creating indices

To create the indices, I first did the imputations in the data for four sensitive questions where there was the option to answer “I do not want to answer this question. Next one!” with the help of the MICE package. Overall, I imputed 48 responses treating such answers as empty cells. Then I conducted Spearman's correlation tests in order to check if variables can be combined into an index as measuring the same constructs, according to theory and previous works.

The variables for the multiplayer gameplay index, the frequency of playing MMORPGs per week and the average collective playing session length, have a statistically significant moderate positive correlation coefficient of rs = 0.462***. All indices were created with the help of the average values.

Table 1 shows Spearman's correlation coefficients for the variables which measured gaming social capital. All the coefficients are significant and positive; the strongest correlation coefficient (rs = 0.553***) can be observed between variables “I feel close to the people I play MMORPGs with” and “I have offline relationships with the people I play MMORPGs with”; the weakest coefficient (rs = 0.187***) is between the sense of belonging to the gaming community and feeling of loneliness without playing MMORPGs.

Table 1. Spearman's correlation coefficients for gaming social capital index variables.

For the interest in politics index, the correlation coefficients between variables are statistically significant and moderately positive (Table 2). The strongest correlation is between the degree of interest in news about politics and the frequency of discussion of political and social news with others equals to rs = 0.603***.



Table 2. Spearman's correlation coefficients for interest in politics index variables.

The visualization of the polychoric correlation coefficients of the civic engagement variables is presented in Figure 12. Civic engagement indicators were divided into three indices according to the correlation matrix. All indices were ranging from 0 to 1. I combined the “vote” and “volunteer” variables into one summary index (rho = 0.840 which is strong positive correlation coefficient); variables signing on paper petitions and online petitions into the second summary index (rho = 0.434 which is moderate positive correlation coefficient) and variables connected to the involvement in some forms of expression of public opinion such as rally, protest, demonstration, and others into the third summary index (rho = 0.688 which is moderate positive correlation). Unexpectedly, the variable on money charity (“Did you donate money to charity over the last three years?”) has a negative correlation with voting (rho = -0.183), volunteering (rho = -0.142), signing paper petitions (rho = -0.107), online petitions (r = -0.059), and permitted rallies (r = -0.081), but it has a weak positive correlation with taking part in “unauthorized rallies” (rho = 0.118); I used this charity variable as a separate one and did not combine it in any index.

The non-civic engagement variables are not correlated with each other, the correlation coefficient is insignificant and weak (r = 0.058) that is why I did not combine them into one index (Figure 13). These two variables, professional/work engagement and participation in different festivals/contests, are not connected to each other. Moreover, I have a variable which measures respondents' incentive to participate deeper in solving problems related to environmental issues, domestic violence, unemployment, drug addiction, etc. All in all, I use these three variables separately. social capital game time

6. Results

In the first regression model (Table 3), I regressed gaming social capital on the multiplayer gameplay and on control variables. The adjusted R-squared of the model with all mentioned above predictors equals 0.432. Multiplayer gameplay index was significant in this model (?? = 0.59, p < 0.001), thus, Hypothesis 1 which posed that multiplayer gaming is positively associated with gaming social capital of players is supported. The sex, city of residence, most-played MMORPG and gaming experience in years are insignificant. However, the frequency of MMORPG play (in hours per week) is a significant control variable which has a concave relationship with a significant negative coefficient for playing any longer than the reference of “1-5 hours” and non-significant coefficient for playing more than 20 hours a week.

Also, age group is a significant variable; there are positive coefficient for age groups “26-30 years old”, “36-40 years old” and “45+ years old” which are compared with the reference category “21-25 years old”.



Table 3. Regression model predicting the gaming social capital index.

Hypothesis 2 predicted that gaming social capital is positively associated with civic actions performed outside the game environment. For this hypothesis, I have four different regression models (see Table 4). In the first model, the vote and volunteering index was regressed on the gaming social capital, interest in politics and control variables. The adjusted R-squared of this model equals 0.409, so the data explains 41% of the variance. The predictors which are significant in the first model (Table 4) are gaming social capital which has a positive association with vote and volunteering index (?? = 0.51, p < 0.001), attitude towards the government actions during COVID-19 virus (?? = -0.18, p < 0.001) and the changes in communication with players in MMORPGs for those who started to communicate more and more closely (?? = -0.22, p < 0.05). In the second model (Table 4) the adjusted R-squared is 0.138; as for the predictors, gaming social capital is again positively related to the offline and online petitions index (?? = 0.18, p < 0.05) and interest in politics also has a positive association with the petitions index (?? = 0.19, p < 0.05). All the other predictors are not significant or only “marginally significant”. In the third model (Table 4), I regressed the rallies index (a combination of attending permitted and unauthorized rallies) on gaming social capital and controls. The model explains only about = 4.7% and there is only one significant predictor, the interest in politics (?? = 0.20, p < 0.05). A similar situation is with the fourth model (Table 4) where I estimated participation in charity (= 8.2%). Neither gaming social capital nor interest in politics are significant predictors in that model, but participants over 45 years old are more likely to do charity as compared to the dummy variable “21-25 years old”. Also, the COVID-19 variable measuring the changes in communication with friends in games is significant (?? = 0.26, p < 0.05). These regression models partly support Hypothesis 2 as only in two out of four models the gaming social capital was a significant variable with positive coefficient. Thus, this hypothesis worked only with such components of civic involvement as voting and volunteering and writing paper and online petitions. Taking part in rallies or donating to charities have other predictors in the context of Russian MMORPG players.

Table 4. Regression models predicting the four civic involvement indices.

Hypothesis 3 states that playing MMORPGs may translate into non-civic engagement. For non-civic engagement, I have three variables and three models (Table 5). The model predicting professional or work engagement (“Have you been an organizer of professional or recreational events during your studies and/or at work over the last three years?”) has two significant positive predictors: gaming social capital (?? = 0.22, p < 0.05) and interest in politics (?? = 0.17, p < 0.05). The adjusted R-squared for this model equals 0.102. The next model predicted the participation in the contests or festivals (Table 5) in the last three years (= 3.6%). The gaming social capital is not a significant predictor here, but the younger the player, the smaller their chances to demonstrate this type of engagement, compared to the age group “21-25 years old”. Also, players who spend most of their time in The World of Warcraft or The Elder Scrolls Online have higher chances for participating in contests or festivals, compared to the other MMORPGs. As for the last model (Table 5) which predicted the incentive to take part in solving social problems (= 23%), those with higher gaming social capital (?? = 0.17, p < 0.05), higher interest in politics (?? = 0.28, p < 0.001), and respondents who aged 31-35 years old (?? = 0.13, p < 0.05), 36-40 years old (?? = 0.16, p < 0.05) and more than 45 years old (?? = 0.14, p < 0.05) compared those from 21 to 25 years old and who have gaming experience more than 10 years (?? = 0.16, p < 0.05) were more likely to tick this answer. Moreover, the sex and years of gaming experience (for category “more than 10 years” compared to “3-4” years) were significant predictors. Notably, the intercept is marginally significant here. These results can partly support Hypothesis 3 because gaming social capital variable was significant and had a positive coefficient in two cases out of three (predicting engagement in work or studies and in participating in solving social problems).



Table 5. Regression models predicting the three non-civic involvement variables.

Hypothesis 4 predicted that excessive gameplay (more than 20 hours per week) is negatively associated with civic engagement; this hypothesis has not been supported. There are 52 respondents in the sample who play excessively. This category is not significantly different from the other in predicting civic engagement; on the contrary, it has a significant positive ??-coefficient in predicting willingness to solve social problems.\



Discussion

The aim of this study was to figure out whether there is a positive association between gaming social capital and civic involvement and to test the hypothesis saying that Massively Multiplayer Online Role-Playing Games can be a tool for creating social ties and gaming social capital which can be then transformed in different kinds of social engagement in real players' life.

Online games are surrounded with controversy and have a mixed reputation among researchers (Anderson and Bushman 2001; Ng and Wiemer-Hastings 2005; Hollingdale and Greitemeyer 2014), but some authors argue that multiplayer video games can bring about social effects that are necessary for community formation and bonding (Kahne, Middaugh, and Evans 2009).

The regression analyses based on an online survey of experienced Russian MMORPG players show that multiplayer gameplay is positively related to higher gaming social capital. The results also indicate that gaming social capital can be associated with civic engagement (voting and signing petitions). This matches the literature saying that, while players interact within the game, they develop social and civic attitudes and then apply them in real life (Molyneux et al. 2015; and Zhong 2011). The sense of belonging to the community as an aspect of gaming social capital was an important measure which was also mentioned in previous research; people who feel connected to other members of the community are more likely to participate in civic and professional social activities.

The results challenge the time displacement theory saying that playing online games, including MMORPG, have an adversary impact on the social life of players. According to the survey, those players who spend over 20 hours per week playing MMORPGs are in most cases not significantly different from other players. There are two exceptions. Participants spending the least and most time playing per week had smaller rates of gaming social capital as compared to those players who spent 6-19 hours per week playing. This could hint towards the idea that there is some “happy mean” in spending time playing where social incentives of the gameplay are positively affecting the social and organizing skills of players but this effect is not observable for beginning/very moderate players and for excessive players. This survey cannot provide a more detailed answer to this question so it should be left to future studies for testing.

All in all, this study contributes to the literature on the social impact of online video games in two ways. First, it presents evidence showing that higher gaming social capital is positively related with voting, volunteering and signing petitions, controlling for interest in politics and other important covariates. Players with higher gaming capital are also more likely to engage in professional activities. This casts a more positive light on the experience gained by players of MMORPGs. Surely, this conclusion needs cannot be overgeneralized and needs further tests on the players of other big game genres such as multiplayer online battle arenas, e.g., Dota2. This conclusion is especially important given the high rates of gaming in the population of Russia.

Second, this study brings more light on the population of MMORPG players as a social group in society. By focusing on the social and civic characteristics of this group, this study debunks popular myths about the largely detrimental effects of gaming and puts them in a data-based context. Moreover, it demonstrates that online gaming environments can stimulate pro-civic people's behavior. This preliminary conclusion can be further tested by using a longitudinal design or matching techniques.