Sexual harassment in online videogames: What we found so far

Sexual harassment is an important issue besetting our society, in particular to how the online and physical worlds are more and more intertwined crossing online platforms and neighborhoods (Citron, 2009; Guardian, 2016; Pew, 2014; Geek feminism timeline of incidents). Nevertheless, there are social scientists, including myself, who examine sexual harassment in that not-so-little online neighborhood called videogames. Thus, I offer a non-exhaustive overview of the sexual harassment literature and what insights researchers have found for online videogames.

There are upsetting language and details in this blog post.

What are the sexually harassing behaviours in videogames?

People were less likely to admit of being sexually harassed, but they were more likely to report experiencing specific sexually harassing behaviours (Magley et al., 1999; Berdahl & Aquino, 2009). Why the difference?  Some behaviours are hard to tell if it's sexually harassing or not, such as sexual jokes, whereas severe behaviours are most obvious to report. Furthermore, they are less likely to report some behaviours when they think it's a normal part of the environment (McCabe & Hardman, 2005).

So we asked 301 gamers to tell us their most recent harassment event, whether they observed or experienced it (Fox & Tang, 2013). We read these descriptions and categorized them. The following is from one player's description of a harassing event:

We found 22 types of behaviours which we made them into a questionnaire format for our second survey. We asked 425 male gamers how often they perpetrated these behaviours and asked 293 female gamers how often they experienced them. Our factor analyses found these behaviours are clustered into two types of harassment: general (i.e., trash talking) and sexual. Female gamers reported a greater variety of harassing behaviours than male gamers reported in perpetrating (Fox & Tang, 2015, 2016; Tang & Fox, 2014, 2016).

The parentheses included an example used in the questionnaire along with which of the genders reported. For example, physical threat was clustered in general harassment for women only.

General Harassment:

*These behaviours have also been reported by Brehm (2013) and Ballard & Welch (2015).

Sexual Harassment:

* These behaviours have also been reported by Brehm (2013), Cote (2015), and Ballard & Welch (2015).

Other behaviours:

*These behaviours have also been reported by Gray (2012).

The other behaviours did not clustered with general, sexual nor as a third cluster. Perhaps these behaviours may be clustered for certain demographics, such as men on the receiving end, ethnic minorities or LGBT players.

How many are affected by sexual harassment in online video games?

The sexual harassment literature found that women experience sexual harassment more often than men do. Pew Research Center (2014) on online harassment found that young women experienced more stalking (9% vs. 6%), and sexual harassment (7% vs. 4%) than men whereas men experienced more name calling (32% vs. 22%), embarrassment (24% vs. 20%), and physical threats (10% vs. 6%). The picture is less clear for online videogames as Pew reported general harassment in which male players experienced it more often.

In our surveys, players reported harassment behaviours on a 5-point scale (never (1) to always (5)) in an online videogame they regularly play. We found that male players perpetrated general harassment in the 'rarely-sometimes' range (M = 2.27, SD = .91) whereas sexually harassed in the 'never-rarely' range (M = 1.23, SD = .56; Tang & Fox, 2014, 2016). On the other hand, female players experienced general harassment (M = 2.98, SD = .87) and sexual harassment (M = 2.77, SD = .99) in the 'rarely to most of the time' range (Fox & Tang, 2015, 2016).

An online survey of 151 players found that players experienced cyberbullying (e.g., name-calling, threats, teased, lies, profanity, hostility) in the 'once or twice in the past 2-3 months' range. Women experienced more sexual harassment, sexual pursuit, group exclusion and being kicked out of their group than men experienced. Furthermore, LGBT players experienced more lies, sexual harassment, sexual pursuit and group exclusion than heterosexual players do (Ballard & Welch, 2015).

A survey of 216 Second Life users found that female players experienced more sexual harassment than male players (38% vs. 13%). Female avatars experienced more sexual harassment than male avatars do. Further analysis revealed that female players with highly sexualised female avatars experienced more sexual harassment, name calling and obscene comments, in contrast highly sexualised male avatars do not experience them (Behm-Morawitz & Schipper, 2015).

Field experiments where researchers play either as a male or female player corroborate survey results. Researchers who posed as a female player received three times as many negative comments as male or silent players (Kuznekoff & Rose, 2013). Further analyses found that other male players who performed worse than a female player tend to utter more negative comments to her  (Kasumovic & Kuznekoff, 2015). Combined with our survey findings, it seems that people with certain predispositions sexually harass women under certain conditions.

Factors of sexual harassment

The sexual harassment literature have identified many factors of why people sexually harass women (Willness et al., 2007). These factors are broadly categorized under person and situation factors.

            Person factors

Most people think that harassment in online games is the exclusive domain of young boys, but our findings say that isn't so (Tang & Fox, 2014). The sexual harassment literature could not identify typical sexual harassment profiles, other than being a man. On the other hand, the literature identified personality traits predicting sexual harassment and here are some of them:

Our first survey examined personality traits related to videogame sexism (Fox & Tang, 2013, 2014). We developed a set of 16 statements reflecting sexist beliefs about videogames (e.g., "Women who call themselves gamer girls think they deserve special treatment") from experts, scholars and female gamers (Fox & Tang, 2014). We found three predictors of videogames sexist beliefs: two aspects of conformity to masculine norms: Power over women (e.g., "I love it when men are in charge of women") and heterosexual self-presentation (e.g., "I would be furious if someone thought I was gay"), and social dominance orientation.

Our second survey extended these findings to harassment behaviours (Tang & Fox, 2016). We found that both hostile sexism and social dominance orientation are significant predictors of sexual harassment. The same is true for general harassment in addition to game involvement and time spent playing videogames. These videogame variables are likely due to being socialized by other gamers, well you learn how to insult by observing others.

So far, our surveys have supported past findings in the sexual harassment literature. Other factors we have yet to explore is whether the gamer identity play a role in sexual harassment, given how women were less inclined to identify as a gamer (Pew, 2015; Shaw, 2012) I should point out that this section focused on the harassers, I have not examine how targets of sexual harassment might be more or less vulnerable to sexual harassment.

            Situation factors

The social environment play a role in setting the tone towards sexual harassment. The literature identified two broad situational factors: gender-context and organizational climate (Willness et al., 2007).

Gender context. The sexual harassment literature found that places where men outnumber women, the women experienced more sexual harassment than those in less male-dominated environments (Willness et al., 2007). When women are few, this puts a psychological spotlight on their gender leading to greater sexualised interactions from their male coworkers. Statistics by the ESA (2015) and Pew (2015) reported a roughly equal gender ratio of people playing videogames, but this is scratching the surface. The gender inequality varies when it comes to specific games. For example, 4.1% reported being female in a League of Legends survey (Ratan et al., 2015); 41.8% among World of Warcraft players (Brehm, 2013); 68% among Second Life users; 25% in our surveys. Thus, we'd expect a higher rate of sexual harassment in male-dominated videogames.

When something is traditionally done by men, seeing women doing the same thing can be jarring for the men. Some men see women as intruders and feel that their masculinity is threatened among other things (Berdahl, 2007). So, one of their reactions is sexual harassment. When people think that videogames is played mostly by men, well they'd tend to sexually harass female players (Taylor, 2006). Some games are more men-only than others, some genres like first-person shooters, fighting, action adventure, sports and strategy gamers are more macho than others, so it's reasonable that there would be more sexual harassment in those games (Cruea & Park, 2012; Eden et al., 2010; Pew, 2015; Phan et al.,  2012). On the other hand, differences could arise from games in the same genre. The literature found that companies emphasizing work over personal obligations reported greater sexual harassment rates than those that balanced them (Timmerman & Bajema, 1999). Perhaps this translates to videogames with high stakes (e.g., ranking, scores, win-loss ratio) reporting higher rates of sexual harassment than games with greater social experiences. For example, Counter-Strike: Global Offensive  vs. Team Fortress 2.

Workplaces with a lot of sexual banter, comments about women's appearances experienced more sexual harassment than those that do not (Berdahl & Aquino, 2009; Timmerman & Bajema, 1999). Given how female videogame characters are frequently hypersexualised, its links with sexual harassment is not farfetched. Studies found that highly sexualised female avatars experienced more sexual harassment than less sexualised female avatars (Behm-Morawitz & Schipper, 2015); sexualised female avatars interviewing for a job were seen as less credible and less trustworthy (Nowak et al., 2015); people who watched sexist media (e.g., sexist posters, sexist TV programs) were more likely to sexually harass and give sexist evaluations (Galdi et al.,  2014; Hitlan et al.,  2009; Mitchell et al.,  2004; Dill et al., 2008).

The organizational climate involve the social norms of how companies and the people within deal with sexual harassment  (Timmerman & Bajema, 1999; Willness et al., 2007). In videogames, this meant how game developers, administrators, moderators, guild/clans leaders among others deal with sexual harassment.

There are three organizational climate aspects that impact sexual harassment (Bergman et al., 2002; Hulin et al., 1996):

The punishments include warnings, bans, shadow bans, suspensions, restrictions (e.g., chat). How some punishments are meaningful across platforms (e.g., Xbox, Steam, Playstation) or games (e.g., DoTA 2, League of Legends) is not reported, with the exception for League of Legends (if there are others I missed, I apologize in advance; Lin, 2015; Lewis-Evans, 2015).

An important situational factor for online sexual harassment is the psychology of internet communication (Barak, 2005). The social identity model of deindividuation effects is one theory that explains how anonymity and identity influence people's online behaviours. First, anonymity does not make anyone equals. Men and women do not interact each other as equals (Postmes & Spears, 2002). For example, women were more likely to hide their gender if they think it puts them at a conversational disadvantage with men, especially when the online neighborhood is male-dominated  (Flanagin et al., 2002). They stereotyped themselves and others with what little knowledge people display online (e.g., redditors, gamers, girlgamers, etc.). Second, prosocial and antisocial online behaviours do not come out of a vacuum. They come out from the online neighborhood's social norms and group membership (Postmes et al, 2001). For example, Starcraft players type "gg" at the end of match as a courtesy. Third, online actions and its consequences are not separate from the physical world. People's real life presence on the Internet is displayed through social media (e.g., Facebook, Twitter, Twitch). People have been physically and psychologically harmed (e.g., doxing, swatting and stalking; Citron, 2009). Thus far, I have only briefly addressed some parts of the psychology of internet communication relevant to online sexual harassment.

Responses to sexual harassment

Pew Research Center (2014) reported that 60% ignored harassment whereas 40% responded to it, the severe harassments were responded to. Those who responded confronted the perpetrators, unfriended or blocked, reported them to the platform, discussed the problem with others, changed or deleted their profile, withdrew, or reported them to law enforcement.

In our survey with female players (Fox & Tang, 2016), we adapted coping strategies from the sexual harassment literature and added strategies related to the Internet, totaling to 31 coping strategies. Our factor analyses found that these coping strategies clustered into five factors, from most frequent to least (5-point scale, never (1) to always (5)):

*These behaviours have also been reported by Brehm (2013) and Cote (2015).

Cote (2015) reported strategies we did not cover in our survey. Female players preferred playing with friends than strangers. Blocking and muting were used occasionally, but inconvenient when teamwork is required. They acted in a more aggressive personality, although it can backfire on them (Berdahl, 2007). Female players did not flirt with other male players because they think it perpetuated the girl gamer stereotype.

Another coping strategy is showing off their gaming prowess. For our survey, this responds to male players doubting female players' motivations and/or abilities for playing video games because of their gender. The downside is that they have to keep their skills up to prevent harassment and it may not work with everyone (Kasumovic & Kuznekoff, 2015). I recognize this coping response to a psychological phenomenon called stereotype threat. Female players harassed about their motivations or abilities face psychological pressures in constantly proving to others which can be tiring (Hall et al., 2015; Kaskan & Ho, 2016). Furthermore, female players underestimate their gaming abilities compared to male players because of the girl gamer stereotype threat (Kaye & Pennington, 2016; Vermeulen et al., 2016; Ratan et al., 2015).

            Confronting perpetrators is not easy

How likely people react with which strategies depend on the severity of the harassment with person and situation factors (Bergman et al., 2002). I'd like to focus on what prevents people from confronting harassers. In the heat of the moment, people are afraid of being called over-sensitive whiners or professional victim (Magley et al., 1999). For male bystanders, they fear of being called a "white knight", homosexual, or emasculated among others (Dickter et al., 2012; Shelton & Stewart, 2004). However, confronters should realize that they felt better about themselves, others (besides the harassers) liked and respected them better and confronters are more effective if they are similar to the harassers (Dodd et al., 2001; Rasinski & Czopp, 2010). In many situations, it helps when you have allies standing up to sexual harassment.

            'Don't feed the trolls' and other common advices on online sexual harassment

The Internet has lot of advices for tackling online sexual harassment. These advices are quite common to have been categorised in the Geek Feminism wiki. These include "toughen up", "laugh it off", "turn the computer off", "report and move on", "don't make a big deal out of it", "called their actions were in good fun" was a behaviour included in our harassment survey. The sexual harassment literature identified these advices as least effective in that they do not communicate the consequences to the harassers (Citron, 2009; Knapp et al., 1997).

These advices have historical roots to domestic violence and sexual harassment in the workplace (Citron, 2009). These advices trivialize online sexual harassment as:

The most effective means to reduce sexual harassment is seeking help from allies, especially from online platforms.

            Organizational response

To my knowledge, there is little academic information regarding online platforms' responses to online sexual harassment. Organizations with clear sexual harassment policies and procedures and who act upon them have lower sexual harassment rates (O'Leary-Kelly et al., 2009). Although, companies have a legal obligation to address sexual harassment among their employees, some online neighborhoods do not have such legal obligations for their users nor do they take it seriously until it is too late (Kessler, 2015; McDonald, 2012).

On the other hand, organizations tackling online sexual harassment face numerous challenges (Matias et al., 2015). Perpetrators may use pseudonymous or multiple accounts. They may group together to harass a single person. Some online videogames may have from a few hundred to several hundred thousand concurrent players, which means in tackling a big volume of harassment reports and false reporting abuses. The large number of reporting would have significant health impact on the human moderators and decrease their effectiveness (Chen, 2014; Lin, 2015; Matias et al., 2015).

Consequences of sexual harassment

When people experience sexual harassment, they suffer physically and psychologically. Physical symptoms include headaches, nausea, respiratory, digestive and sleeping problems among others (Chan et al., 2008; Willness et al., 2007).

Psychological symptoms include anxiety, depression, sadness, lowered self-esteem among others (Chan et al., 2008; Willness et al., 2007). Studies have linked sexual harassment experiences to symptoms of post-traumatic stress disorder (Avina & O'Donohue, 2002). Rumination (e.g. "rehashing their experiences with harassment in their mind") mediates the relationship between sexual harassment experiences and its consequences (Fox & Tang, 2016; Runions et al., 2013). Because sexual harassment focused on women's bodies, it is linked to women's sexual self-objectification which have similar psychological consequences (Davidson et al., 2015; Fairchild & Rudman, 2008).

Sexual harassment can create a stressful environment to people who are indirectly exposed, this is called ambient sexual harassment (Glomb et al., 1997). Bystanders are worried that they may be targeted next and such hostile environment resulted in similar consequences on their physical and psychological health (Miner-Rubino & Cortina, 2004; Hitlan et al., 2006).

The consequences of sexual harassment in the workplace is translatable to online videogames (Lapierre et al., 2005). First, players would find the online videogame and its players less enjoyable. Second, players would spend less time, money and effort in that online videogame given its toxicity. Third, players' performance might drop, an infamous example is the Cross Assault harassment incident (video).

In our survey with female players (Fox & Tang, 2016), we found that sexual harassment have direct and indirect effects to withdrawal from gaming. As they experience more sexual harassment, they are more likely to withdraw from gaming (e.g., "If I'm being harassed, I quit in the middle of a game"), they ruminate more, and their perception of organizational responsiveness to harassment decreases (e.g., "To my knowledge, they [administrators, moderators, etc.] investigate harassment complaints no matter what type of harassment it is"). The rumination and organizational responsiveness also affect female players' likelihood in withdrawal from gaming. The relationship for general harassment only showed a direct relationship to withdrawal from gaming. We noted that female players pre-emptively avoided experiencing sexual harassment through gender masking, but are still affecting by being bystanders of others' sexual harassment.

A potential consequence is people's identity as gamers (Grooten & Kowert, 2015). According to Pew (2015), fewer women self-identify as gamers than men do. Although there are a variety of reasons for the discrepancy, sexual harassment may be one of them. The hostility towards female gamers is a form of rejection which weaken their gamer self-perception. Furthermore, sexual harassment may reduce their gaming creds as they are pushed off from participating in forums, etc. However, this does not mean that they will stop playing, but rather disassociate their gaming from the gamer identity. This disassociation may include avoiding the social aspects of gaming, masking their gender and keeping their gaming activity to themselves. For bystanders, they may disassociate their gamer identity when it becomes associated with undesirable cultural attributes, such as misogyny.

Future research directions - We're going need a bigger boat

I have covered the sexual harassment literature extensively with related videogames studies. But, this overview is not exhaustive as there are many issues needing attention and I can do only so much on my own.

Across the sexual harassment videogames studies, the majority of survey participants were white/Caucasian making analyses on racism in online videogames impossible. To my knowledge, Dr. Kishonna Gray (Eastern Kentucky University) is among the few who investigate the experiences of women and people of color in the gaming community (Gray, 2012, 2012a, 2013). Another issue is the social interaction between male and female players of different countries and cultures. I am curious how sexual harassment is dealt with in different servers. Some games have one server like EVE Online where players around the world are interacting in the same space, whereas other games have servers for specific geographic regions (e.g., North America, South America, Asia, Europe, Brazil, Russia, etc.).

The majority of sexual harassment targets are women. Nevertheless, there are sexual harassment studies of men against men, women against men, women against women and LGBT individuals (DeSouza & Solberg, 2004; Sibley et al., 2007; Waldo et al., 1998). There are sexual harassment accounts of female videogames developers, although academic studies are few (Hamilton, 2012; Jenson & de Castell, 2013). To my knowledge, there are no studies on live streamers (e.g. Twitch.tv), although it has some news articles (Hernandez, 2013; Grayson, 2015). Notably, people were focused female streamers' appearance and their performance, which relates to research on female athletes and the media (Nezlek et al., 2014).

Most sexually harassing behaviours are aggressive and one of its driving personality trait is hostile sexism. On the other hand, there is a subjectively positive, yet pernicious form of sexism called benevolent sexism (Fiske & Glick, 1995; Fraser, 2015). Benevolent sexism is a view that men should protect and cherish women as wife, mother and caretaker (and virgin) limiting what women can do and be. They harass women who don't turn out to be their ideal image of women. Men with such beliefs are known as "white knights". In videogames, players giving female players unsolicited gifts, romantic affection, and protective companionship among others reflect such benevolent sexism (Linderoth & Öhrn, 2014).

The current climate in the videogame community pose a significant barrier for future sexual harassment research. There is hostility to researchers who study gender issues and skepticism regarding gender discrimination studies, in particular from men and people high in hostile sexism and I would not be surprised to see similar reactions to this piece (Kim & Tidwell, 2014; Moss-Racusin et al., 2015). This gendered skepticism extends to legislative efforts in curbing online harassment (Citron, 2009), such as how a congresswoman was swatted because of her anti-swatting legislative efforts (Machkovech, 2016).

Since the events known as Gamergate, several games scholars were attacked and harassed because of their work on gender issues (Chess et al., 2014; Ivory et al., 2014). People contaminated online surveys because they saw these surveys as undesirable or a threat to their social identity (Nauroth et al., 2014; 2015; Belmi et al., 2015). For example, a comic book researcher and a games researcher directly affected by the Gamergate events were harassed because they touched on gender issues (Allaway, 2014; Khouri, 2014). Luckily, we conducted our surveys long before Gamergate, but future studies will be more challenging as we must carefully guard against malicious responses and threats.

That is all for now.

I’d like to thank Jamie Madigan for his thoughts on an earlier version of this blog post.

References posted at VG Researcher.

Further Reading

Barak, A. (2005). Sexual harassment on the internet. Social Science Computer Review, 23 (1), 77-92. URL http://dx.doi.org/10.1177/0894439304271540

Citron, D. K. (2009). Law's expressive value in combating cyber gender harassment. Michigan Law Review, 108 (3), 373-415. URL http://www.michiganlawreview.org/articles/essay-law-s-expressive-value-in-combating-cyber-gender-harassment

O'Leary-Kelly, A. M., Bowes-Sperry, L., Bates, C. A., & Lean, E. R. (2009). Sexual harassment at work: A decade (plus) of progress. Journal of Management, 35 (3), 503-536. URL http://dx.doi.org/10.1177/0149206308330555

Pina, A., & Gannon, T. A. (2012). An overview of the literature on antecedents, perceptions and behavioural consequences of sexual harassment. Journal of Sexual Aggression, 18 (2), 209-232. URL http://dx.doi.org/10.1080/13552600.2010.501909

Pina, A., Gannon, T. A., & Saunders, B. (2009). An overview of the literature on sexual harassment: Perpetrator, theory, and treatment issues. Aggression and Violent Behavior, 14 (2), 126-138. URL http://dx.doi.org/10.1016/j.avb.2009.01.002