Stanford Prison experiment, move aside. There’s a new test of character in town and its name is cyber-reality. In a new study, researchers discovered that the personality of a video game representation influences a person’s real-world behavior… so much so that those who played a villain were more inclined to act meanly toward others after the game ended. "In virtual environments, people can freely choose avatars that allow them to opt into or opt out of a certain entity, group, or situation," said Gunwoo Yoon, a researcher at Institute of Communications Research, University of Illinois at Urbana-Champaign. "Consumers and practitioners should remember that powerful imitative effects can occur when people put on virtual masks."


A team of two researchers wished to investigate “whether certain types of avatars and avatars’ behaviors could promote pro- or antisocial actions in everyday behavior.” In other words, would those who chose a heroic avatar behave honorably? Alternately, would those who chose a villainous avatar behave dishonorably?

For their study, they enlisted the help of 194 participants: 95 male, 99 female, with an average age of about 20. Participants were told they would be involved in two separate studies. The researchers randomly assigned participants to the avatar of Superman (a hero), Voldemort (a villain), or a circle (neutral). The participants played for five minutes and, as their avatars, they fought off enemies. When that ended, participants moved on to an “unrelated” study, described as a blind taste test. After sampling chocolate and chili sauce, the participants were instructed to give one or the other to a future participant, by pouring the chosen food into a plastic dish. They understood that the future participant would be required to consume all of the food provided.

Have you already guessed the results of the study? The Superman participants poured, on average, nearly twice as much chocolate as chili sauce for the pretend "future participant," and they poured significantly more chocolate than those who played as either of the other avatars. Comparatively, the Voldemort participants poured out nearly twice as much chili sauce than chocolate, and they poured significantly more chili sauce in comparison to the other participants. The researchers confirmed these results with an additional 125 undergraduate participants and this time, they also discovered playing the video game produced stronger effects on subsequent behavior than just watching someone else play. “Our results indicate that just five minutes of role-play in virtual environments as either a hero or villain can easily cause people to reward or punish anonymous strangers," Yoon stated in a press release.

Some might argue that five minutes of pretending to be Voldemort simply increases your desire for spicy food, but the fact is the level of identification hardly mattered at all to how participants behaved; whether the gamers reported themselves as barely or substantially identifying with their avatar, their behavior showed some influence of the on-screen alter ego. "People are prone to be unaware of the influence of their virtual representations on their behavioral responses," the authors noted. Meanwhile, another set of researchers questioned how much our real lives influence our virtual lives.

Stereotype Lift

As every gamer knows, self-representation in virtual environments may have little to no reference to real world identity. Thin people in real life may appear massive on-screen, while others have a penchant for gender-bending; a petite women may choose a warrior to represent herself to opponents. Even better, many avatar-based environments provide users with customization options so gamers can tweak their image to their heart’s content. A team of researchers from the Ohio and Stanford universities explored, in their own words, “how social category representation via avatars (i.e., graphical representations of people in computer-mediated environments) affects stereotype-relevant task performance.” In particular, they wanted to understand whether and how a performance boost occurred when avatar-based gender representation was arbitrary. They were investigating stereotype lift in the virtual world.

Stereotype lift is a term for the performance boost that occurs when a person is aware that another group of people — a group he or she does not identify with — is negatively stereotyped. Commonly, for instance, women are thought of as less talented in the areas of math and science, while men are often thought less capable of empathy. In their experiment, then, the researchers hypothesized that male avatars competing against female avatars would display a stereotype lift when performing a task normally associated with poor performance on the part of females. Their theory was based in part on past studies that have shown how gender-based perceptions are likely to be greater in mixed-gender groups, as compared to same-gender groups — especially when groups are formed ad hoc.

For the experiment, then, the researchers randomly assigned 120 female and male participants either a female or a male avatar. With respect to their on-screen gender, participants believed they were the only virtual female (or virtual male) within their threesome, arranged by the researchers. Next, participants were instructed to perform an arithmetic task, either competitively or cooperatively, with their two opposite-gender avatar companions. What did the researchers discover?

Participants represented by a male avatar and competing against two female avatars showed the strongest performance in comparison to any other condition (such as, a male working cooperatively with two females or a female competing against two males). Painfully, this pattern occurred regardless of a participant's actual gender, suggesting a virtual stereotype lift influenced their behavior. In other words, the participants did not escape stereotypes, created and deeply reinforced, when they entered a virtual world, one in which they could potentially be more free.

Translated from the Sanskrit, avatar means "incarnation," it is the human or animal form of a Hindu god on earth. If our virtual incarnations can influence our behavior even to the smallest extent, maybe there is great wisdom to be gained by listening to the many who believe the spiritual world far outweighs the physical.


Yoon G, Vargas PT. Know Thy Avatar: The Unintended Effect of Virtual-Self Representation on Behavior. Psychological Science. 2014.

Lee JE, Nass CI, Bailenson JN. Does the Mask Govern the Mind?: Effects of Arbitrary Gender Representation on Quantitative Task Performance in Avatar-Represented Virtual Groups. Cyberpsychol Behav Soc Netw. 2014.