An exploration of research on avatars, possible selves, and how game design can positively impact players. (2016)
One of the most promising areas for embedding psychological interventions in interactive media is via avatars. When we play a game with an avatar, we are roleplaying. Our sense of self significantly merges with the avatar, as we adopt the character’s perspective to a greater degree than in other media. Research demonstrates that qualities of avatars like their actions, appearance, and personality can influence the way players interact with virtual spaces and even how they think or behave after playing. These studies reveal the potential of avatars as key mechanisms for self-awareness, growth, and positive behavior change. Specifically, avatars could function in a game to support players’ psychological well-being. I will suggest two possible areas for intervention using avatars. One is identifying and cultivating positive traits (character strengths); another is imagining ideal possible selves. Both are based on interventions shown to influence well-being variables like health, satisfaction with life, and depression. Then I will discuss the evidence about how players commonly use avatars and the ways avatars can influence players, in support of those suggestions.
Character Strengths
Character strengths are positive traits that are relatively stable and contribute to fulfillment. Peterson and Seligman (2004) identified twenty-four character strengths that are valued by all cultures, including things like creativity, prudence, honesty, appreciation of beauty, and curiosity. Character strengths are malleable, and can be intentionally cultivated through practice. All people have all of these strengths to one degree or another, but every person has a handful that are particularly important to their lives and their sense of self, known as “signature” character strengths (Peterson & Seligman, 2004). Using one’s signature strengths has been found to improve well-being. In one study, participants took an inventory that ranked them on all twenty-four character strengths, and were instructed to use their top five in a new way every day for a week. This exercise decreased their depression and increased their happiness for up to six months (Seligman, Steen, Park, & Peterson, 2005). Other studies have replicated similar findings for this activity (ex., Gander, Proyer, Ruch, & Wyss, 2013).
Because character strengths are similar to both traits and skills, they could be easily integrated into the personality or attributes of an avatar. If character strengths were made to be an important feature of avatars—for instance, like customizable traits in The Sims or abilities in RPG skill trees—they could provide a new dimension on which players identify themselves in their avatars, or represent how they would like to be. By adopting the roles of characters with particular strengths, players could have the opportunity to recognize their own positive qualities and values, and to practice expressing the characteristics they want to cultivate.
Of course, many games already depict character strengths, especially, for example: bravery, kindness, leadership, persistence, and fairness. Using the character strengths classification as a source for design ideas could help those positive qualities become more salient to players. This framework also raises questions about “misuse” or overuse of character strengths that we see in games: for instance, attributes similar to the strength of social intelligence (like charisma) often allow characters to selfishly manipulate others, with little consequence. Thinking about charisma in a different way could give rise to new mechanics. Additionally, the strengths lens could call attention to qualities that are less often explored in interactive media. How might a character behave, for instance, if they were high in strengths of forgiveness, or modesty, or gratitude? Designers are developing strategies for considering, intentionally promoting, and evaluating the value content in games (Flanagan & Nissenbaum, 2014), which would be highly relevant to integrating character strengths and virtues.
Best Possible Selves
Another potential source for avatar-based interventions is the best possible self exercise. In the original study, participants responded to a writing prompt once every day for four days. They were instructed to free-write about their future lives, assuming everything had gone as well as possible and they had met all their goals, and to describe that future in great detail (King, 2001). Compared to a neutral prompt, writing about goals in this way was associated with fewer health center visits and better mood. The exercise also increased subjective well-being (including satisfaction with life), a benefit which persisted to a small degree even three weeks later. Similar results on mood have been found in more recent studies on this exercise (Sheldon & Lyubomirsky, 2006) and similar writing prompts have been shown to increase optimism independently of their influence on positive emotions (Meevissen, Peters, & Alberts, 2011; Peters, Flink, Boersma, & Linton, 2010).
An avatar could readily function as a vessel for imagining and enacting a best possible future self. Researchers have reasoned that the intervention’s positive effects are due to the narrative structure of written responses and the high level of detail (King, 2001); it’s possible those key conditions could be met in a game context. Furthermore, since health benefits are found even when participants write about other people’s experiences (as referenced in King, 2001), the intervention might be effectively integrated without avatars necessarily “sharing” players’ real-life goals. For some outcomes, like optimism, it may be sufficient for players to customize a highly detailed fictional “best possible future” for their avatars. Doing so would help the intervention be integrated into a game without making it feel too unfamiliar. This strategy would be in line with the Embedded Design approach, which describes ways to “stealthily” incorporate persuasive content into games. Preliminary evidence for the Embedded Design approach shows that such covert strategies can make games for change more effective (Kaufman & Flanagan, 2015).
How Digital Avatars Influence Players
The above suggestions are supported by bodies of research about how avatars can constructively influence players. Specifically, those mechanisms of influence include: the Proteus effect, customization, identification with a self-similar avatar, wanting to be like idealized avatars, mediated experiences of self-efficacy, and generating possible selves.
Using avatars that look a certain way or function in a particular role can influence us for the better. This is, in part, supported by research on the Proteus effect: the unconscious process by which visual cues from virtual representations inform player behaviors to be consistent with that representation (Yee & Bailenson, 2007). In other words, you make decisions in a game partially because of what you think your avatar’s appearance says about how “you” would behave. Since interactive media is more involving than media that’s passively consumed (Cole, Yoo, & Knutson, 2012), the Proteus effect is generally found to be stronger than when the same concepts or character traits are simply primed, or brought to mind (Yee & Bailenson, 2009). Therefore, if avatars in a game are perceived to be prone to certain constructive attitudes and behaviors (perhaps because she behaves heroically, or simply because he’s wearing white instead of black), players are more likely to adopt those constructive qualities while playing (like helping other players or characters). This is especially useful to know if an intervention is embedded into gameplay itself: for instance, in “exergames” that use motion detectors for game controls, encouraging physical exertion. In one study, participants assigned to use an avatar with a normal weight in a Wii tennis game were more active than those assigned to an otherwise identical avatar with an obese-range BMI, regardless of players’ actual weight. This supports the idea that players act in accordance with implicit expectations of their avatars; in this case, regarding athletic skill (Peña & Kim, 2014).
Avatar effects can also persist after the game is over. For instance, there is evidence that playing a helping role in a game can make players more likely to engage in prosocial behavior immediately after playing. Some prosocial behaviors that have been studied include helping experimenters pick up materials they had “accidentally” spilled and speaking up when an experimenter was apparently being harassed (Greitemeyer & Osswald, 2010). Evidence on the Proteus effect and persisting avatar effects supports the idea that using avatars with salient character strengths could nudge players towards recognizing and practicing those strengths in real life.
Customization and Identification
Customization intensifies the effect of avatars because it makes us more likely to identify with the avatar in some way. Identification is a process by which people adjust their self-concepts to match a character while engaging with media; the activation of those concepts can persist in the time immediately following engagement (Klimmt, Hefner, Vorderer, Roth, & Blake, 2010). Studies have demonstrated that when customizing their avatars, players typically use strategies that make the avatar easier to identify with. In one study that manipulated whether participants customized their game avatars or not, customization was positively related to different kinds of identification with the avatar—notably, both similarity identification (feeling that the avatar is like you) and wishful identification (wanting to be like the avatar). In turn, all types of identification were associated with intrinsic motivation to play the game and with positive mood. The stronger the identification, the better the outcome on those measures (Birk, Atkins, Bowey, & Mandryk, 2016). This suggests that, given the chance, people generally prefer avatars that are both like themselves and like an ideal version of themselves, and that making those choices leads them to enjoy the game more. This conclusion is also supported by survey data in which MMO players self-report both of these motivations as among the most common for their character customizations (Ducheneaut, Wen, Yee, & Wadley, 2009). There is also some evidence that on the whole, players create avatars to be similar to their ideal selves more than their actual selves. This was true in a survey of World of Warcraft players for certain personality traits (Bessière, Seay, & Kiesler, 2007) and in a study of adult women who created Second Life avatars that reflected their ideal body silhouettes (Thomas & Johansen, 2012). It’s important to note that people hold their own individual ideals; they are not wholly determined by social desirability. Also, preferences for how closely players wish avatars to resemble their actual selves can depend on a number of factors, including their age (Ducheneaut et al., 2009), their satisfaction with life (Trepte & Reinecke, 2010) and how depressed they are (Bessière et al., 2007). Since many players, at least in some contexts, view their avatars as having traits they would like to have in real life (Bessière et al., 2007), it is reasonable that this would also apply to character strengths.
Identification and Behavior Change: The Role of Self-Efficacy and Possible Selves
When it is easy to identify with a media character, it is more likely that the character can serve as a behavior model to build a player’s self-efficacy. Self-efficacy is a domain-specific belief in one’s ability to successfully carry out some action or complete some task. Self-efficacy is important for things like goal-striving and persistence in the face of setbacks, and is predictive of successful behavior change (Bandura, 1982). We can see this at work in video game research: for instance, cancer-related self-efficacy (in combination with cancer-related knowledge) was the mediating factor in the success of the game Re-Mission for improving treatment adherence in young people with cancer (Kato, Cole, Bradlyn, & Pollock, 2008). Since self-efficacy helps build mastery, it is also related to competence, an important aspect of psychological wellness (ex., Ryan, Huta, & Deci, 2008). The most effective way to develop self-efficacy is through first-hand experience. It is also possible to build self-efficacy by “vicarious induction” (Bandura, 1982, p. 126): by observing someone else successfully handle a situation. In this way, people can learn effective strategies to deal with the observed situation; understand how predictable or controllable that situation is; and draw inferences about their own ability to deal with that situation. But the type of behavior model matters: the more the model is perceived to be similar to the self, the more self-efficacy will be built by vicarious induction (Bandura, 1982). Video games offer something in between first-hand and vicarious experience, called “mediated enactive experience” (Peng, 2008, p. 649). In an experiment to demonstrate this, participants who played a game about making healthy food choices reported greater diet-related self-efficacy than those who only watched a video of the game. This effect was mediated by identification: those who played the game identified more with the game’s character than those who watched, and this factor drove the differences in self-efficacy (Peng, 2008).
Interactivity alone does not automatically produce strong identification with a character, however. The degree of identification a player experiences with their avatar can vary based on similarity to the self (Trepte & Reinecke, 2010). This means that if an avatar is similar to a player in some way—even if it was not customized—processes like the Proteus effect are likely to be stronger than if the avatar is dissimilar.
The similarity effect is strongest, of course, when it’s really you. Digital avatars make this easy to achieve. In one study, experimenters made digital models from 3D scans of participants’ heads. Then participants viewed a virtual scene in which a digital character first encountered a frustrating situation, and then used a deep breathing exercise to calm down. Half of participants viewed characters that were the representations of themselves, and the other half saw a dissimilar character. Even though this scenario was not really interactive, EEG data suggest that participants experienced similar arousal states in synchrony with the character, but only if it was a self representation (Wrzesien et al., 2015). A different experiment, in which subjects exercised in virtual reality, demonstrates that these similarity effects can actually influence behavior (Fox & Bailenson, 2009). Digital photos were used to construct avatars that looked like participants. In virtual reality, avatars were viewed from a third-person perspective so that participants visually recognized the appearance and behavior of their avatars. For some participants, it was a dissimilar looking avatar rather than their self-representation. After an initial physical exercise procedure, in which the avatars worked out in synchrony with the participants, they were allowed to continue exercising if they wanted, for however long they wanted, before ending the experiment. Those who had a self-similar avatar exercised more in this phase than those with an “other” avatar. In a follow-up study, participants saw similar- or different-looking characters in virtual reality running on a treadmill. Those who saw a virtual self then reported exercising more in the following twenty-four hours than other participants. Thus, interactive media for behavior change can be particularly effective when it appears that you yourself are already engaging in the target behavior.
Part of why self-similar models can enhance self-efficacy for behavior change is because they represent possible selves. Possible selves are conceptions of the qualities or roles we might have in the future (Markus & Nurius, 1986). They can be either desirable selves or undesirable, feared selves. Possible self-concepts motivate related behaviors that can move people towards a desired self or away from a feared self. But possible selves don’t appear out of thin air; rather, we derive possible selves from sources like role models, our personal experiences, cultural norms, and media figures. The roles we take on while playing games can be one such source (ex., Beier, Miller, & Wang, 2012). Although it has not yet been studied specifically, I hypothesize that the more we identify with a game avatar, or the more an avatar appears to represent our actual selves, the more likely we are to see that character as a viable possible self in terms of its realistic qualities and contexts.
When it comes to exploring the potential of game avatars as a way to covertly deliver interventions to a wide audience, the ideas presented in this paper are simply the low-hanging fruit. Character strengths and best possible selves are concepts that align neatly with what we know about how avatars function psychologically and what kinds of avatars players prefer to use. Investigating these topics further could be a productive starting point for exciting advancements in both game design and psychology.
Bibliography
Bandura, A. (1982). Self-efficacy mechanism in human agency. American Psychologist, 37(2), 122–147. https://doi.org/10.1037/0003-066X.37.2.122
Beier, M. E., Miller, L. M., & Wang, S. (2012). Science games and the development of scientific possible selves. Cultural Studies of Science Education, 7(4), 963–978. https://doi.org/10.1007/s11422-012-9408-0
Bessière, K., Seay, A. F., & Kiesler, S. (2007). The ideal elf: Identity exploration in World of Warcraft. CyberPsychology & Behavior, 10, 530–535. https://doi.org/10.1089/cpb.2007.9994
Birk, M. V., Atkins, C., Bowey, J. T., & Mandryk, R. L. (2016). Fostering intrinsic motivation through avatar identification in digital games. In Proceedings of the 2016 CHI Conference on Human Factors in Computing Systems (pp. 2982–2995). New York, NY, USA: ACM. https://doi.org/10.1145/2858036.2858062
Cole, S. W., Yoo, D. J., & Knutson, B. (2012). Interactivity and reward-related neural activation during a serious videogame. PLoS ONE, 7(3), e33909. https://doi.org/10.1371/journal.pone.0033909
Ducheneaut, N., Wen, M.-H., Yee, N., & Wadley, G. (2009). Body and mind: A study of avatar personalization in three virtual worlds. In Proceedings of the SIGCHI Conference on Human Factors in Computing Systems (pp. 1151–1160). New York, NY, USA: ACM. https://doi.org/10.1145/1518701.1518877
Flanagan, M., & Nissenbaum, H. (2014). Values at play in digital games. Cambridge, Massachusetts: The MIT Press.
Gander, F., Proyer, R. T., Ruch, W., & Wyss, T. (2013). Strength-based positive interventions: Further evidence for their potential in enhancing well-being and alleviating depression. Journal of Happiness Studies, 14(4), 1241–1259. https://doi.org/10.1007/s10902-012-9380-0
Greitemeyer, T., & Osswald, S. (2010). Effects of prosocial video games on prosocial behavior. Journal of Personality and Social Psychology, 98(2), 211–221. https://doi.org/10.1037/a0016997
Kato, P. M., Cole, S. W., Bradlyn, A. S., & Pollock, B. H. (2008). A video game improves behavioral outcomes in adolescents and young adults with cancer: A randomized trial. Pediatrics, 122(2), e305–e317. https://doi.org/10.1542/peds.2007-3134
Kaufman, G., & Flanagan, M. (2015). A psychologically “embedded” approach to designing games for prosocial causes. Cyberpsychology: Journal of Psychosocial Research on Cyberspace, 9(3). https://doi.org/10.5817/CP2015-3-5
King, L. A. (2001). The health benefits of writing about life goals. Personality and Social Psychology Bulletin, 27(7), 798–807. https://doi.org/10.1177/0146167201277003
Klimmt, C., Hefner, D., Vorderer, P., Roth, C., & Blake, C. (2010). Identification with video game characters as automatic shift of self-perceptions. Media Psychology, 13(4), 323–338. https://doi.org/10.1080/15213269.2010.524911
Markus, H., & Nurius, P. (1986). Possible selves. American Psychologist, 41(9), 954– 969. https://doi.org/10.1037/0003-066X.41.9.954
Meevissen, Y. M. C., Peters, M. L., & Alberts, H. J. E. M. (2011). Become more optimistic by imagining a best possible self: effects of a two week intervention. Journal of Behavior Therapy and Experimental Psychiatry, 42(3), 371–378. https://doi.org/10.1016/j.jbtep.2011.02.012
Peña, J., & Kim, E. (2014). Increasing exergame physical activity through self and opponent avatar appearance. Computers in Human Behavior, 41, 262–267. https://doi.org/10.1016/j.chb.2014.09.038
Peng, W. (2008). The mediational role of identification in the relationship between experience mode and self-efficacy: Enactive role-playing versus passive observation. CyberPsychology & Behavior, 11(6), 649–652. https://doi.org/10.1089/cpb.2007.0229
Peters, M. L., Flink, I. K., Boersma, K., & Linton, S. J. (2010). Manipulating optimism: Can imagining a best possible self be used to increase positive future expectancies? The Journal of Positive Psychology, 5(3), 204–211. https://doi.org/10.1080/17439761003790963
Peterson, C., & Seligman, M. (2004). Character strengths and virtues: A handbook and classification. Washington, DC / New York: American Psychological Association / Oxford University Press.
Ryan, R. M., Huta, V., & Deci, E. L. (2008). Living well: A self-determination theory perspective on eudaimonia. Journal of Happiness Studies, 9(1), 139–170. https://doi.org/10.1007/s10902-006-9023-4
Seligman, M. E. P., Steen, T. A., Park, N., & Peterson, C. (2005). Positive psychology progress: Empirical validation of interventions. The American Psychologist, 60(5), 410–421. https://doi.org/10.1037/0003-066X.60.5.410
Sheldon, K. M., & Lyubomirsky, S. (2006). How to increase and sustain positive emotion: The effects of expressing gratitude and visualizing best possible selves. The Journal of Positive Psychology, 1(2), 73–82. https://doi.org/10.1080/17439760500510676
Thomas, A. G., & Johansen, M. K. (2012). Inside out: Avatars as an indirect measure of ideal body self-presentation in females. Cyberpsychology: Journal of Psychosocial Research on Cyberspace, 6(3). https://doi.org/10.5817/CP2012-3-3
Trepte, S., & Reinecke, L. (2010). Avatar creation and video game enjoyment: Effects of life-satisfaction, game competitiveness, and identification with the avatar. Journal of Media Psychology: Theories, Methods, and Applications, 22(4), 171–184. https://doi.org/10.1027/1864-1105/a000022
Yee, N., & Bailenson, J. (2007). The Proteus effect: The effect of transformed self-representation on behavior. Human Communication Research, 33(3), 271–290. https://doi.org/10.1111/j.1468-2958.2007.00299.x
Yee, N., & Bailenson, J. N. (2009). The difference between being and seeing: The relative contribution of self-perception and priming to behavioral changes via digital self-representation. Media Psychology, 12(2), 195–209. https://doi.org/10.1080/15213260902849943
Sydney Rubin 