Online Video Games, a Mirror of the Personality of Internet Users?
Sometimes, games can tell a lot about the character of the players 1 . For example, one day, the father of one of the authors of this article noticed that he liked to play golf with his business partners, not only to socialize with them, but especially to see who cheats on the course. This implies, of course, that the way someone behaves during a golf course is a reflection of his personality and his business behavior. Moreover, it seems that players do not even have to be face-to-face to assess the personality of their partners. For example, in the case of online video games sites like this http://addictinggamesz.com, we note that players who form romantic relationships in these virtual worlds report a similar phenomenon (Yee, 2005):
“The game is why we fell in love. Going through all these adventures and quests together really built our relationship. We could see how the other behaved when he was angry, tired, sad, happy … “[City of Vilains player, 25 years old].
2This suggests that video games could be an interesting platform for observing the intersection between player personality and online behavior. But this enthusiasm must be tempered: these games are very often fanciful (for example, inspired by the works of John R. Tolkien) and, consequently, they could also encourage behaviors very different from those of everyday life. This tension between the potential of video games as a mirror image of players’ personalities and, conversely, their potential to encourage unusual behavior is at the root of two founding questions of the research project that we will describe here:
31. Is it true that a user’s personality can be detected from his or her behavior in a video game?
42. If so, what indicators are particularly revealing of specific personality traits (such as introversion or openness)?
5Apart from the purely academic interest of these questions, this research area also has an important practical potential. Indeed, researchers in human-machine interaction have long been interested in the idea of personalizing computer systems (Mackay, 1991, Riecken, 2000). Learn more about the personality of users could help create systems more suited to their needs. It may even be possible to adjust the properties of a system over time, as the user’s personality is revealed through the behavioral indicators followed by the software. We will use the data obtained in the popular online video game World of Warcraft ( w o w) to answer the two previous questions, while also reflecting important limitations suggested by our research.
The expression of personality
6Studies in the psychology of personality have repeatedly shown that two strangers meeting for the first time are able to quickly and accurately assess each other’s personality. Crucially, the indicators (gestural, verbal …) used by these strangers are the subject of an almost universal consensus – in other words, it seems that the personality is visible in everyday life through indicators common to all. Several studies show that this is true even in very brief encounters (Funder, Sneed, 1993, Kenny, Horner, Kashy, Chu, 1992). For example, a study using video-recorded face-to-face conversations (Funder, Sneed, 1993) shows that extroverts speak louder, with more enthusiasm and energy, and also that they are more expressive with their gestures. Other studies have sought to predict personality through examination of an individual’s room or office (Gosling, Ko, Mannarelli, Morris, 2002), or even his music collection (Rentfrow, Gosling, 2006) . In studies of personal space, conscientious individuals had well-lit, clean, and well-organized rooms. Similarly, individuals with a high score for open-minded had more books and magazines. This type of research has also been extended to the computer field. In particular, studies have shown that it is possible to form a moderately accurate impression of an individual’s personality based on his website (Marcus, Machilek, Schutz, 2006, Vazire, Gosling, 2004), his profile on Facebook (Back et al., 2010), the content of his email (Gill, Oberlander, Austin, 2006), his blog (Yarkoni, in press), and even simply on the basis of his email address – one of the simplest indicators of the electronic world (Back, Schukle , Egloff, 2008). To illustrate these phenomena, we note for example that the linguistic content of blogs belonging to an individual classified as pleasant contains more positive words (“happy”, “happy” …), more words related to the family, and more. of sentences using the first person. Conscientious individuals tend to use more words that correspond to personal success. These studies show that we leave traces of our personality both in the physical world and in the virtual world. If we consider that, on average,
The limits of the expression of the personality
7In contrast to the previous section, it should also be noted that there is reason to believe that personality does not express itself so clearly in virtual worlds. First, the studies above have largely focused on the expression of personality in a daily, ordinary context. It may not express itself so clearly in a fantastic context, through nonhuman bodies, capable of extraordinary activities (for example, a gnome priest resurrecting the dead with rays of magic light). In connection with this, some researchers like Sherry Turkle (1997) have suggested that virtual worlds, far from reflecting our personality, allow the player to constantly reinvent itself. If we interpreted this concept literally, it follows that there is a clear divide between a player’s personality and his online behavior. In other words, if players can reinvent themselves at will in virtual worlds, there is probably no consensus indicator of online personality. Some data suggest that this might be the case, showing that Internet users indeed change their behavior in virtual worlds. So studies show that “role players” ( showing that Internet users change their behavior in virtual worlds. So studies show that “role players” ( showing that Internet users change their behavior in virtual worlds. So studies show that “role players” (role-player ) are generally more imaginative and able to experiment with their characters (Caroll, Carolin, 1989, Simon, 1987). Other studies of online dating sites (Hancock, Toma, Ellison, 2007) and video games (Bessiere, Seay, Kiesler, 2007) show that users, in both cases, idealize their presentation to varying degrees. variables. In particular, some studies indicate that this trend towards online idealization is moderated by self-esteem – low self-esteem leading to more idealization (Ducheneaut, Wen, Yee, Wadley, 2009). Therefore, it is possible that these identity reinventions and their variations across individuals suppress the expression of stable personality indicators within virtual worlds.
Collection of personality data
8In general, studies on the expression of personality are generally based on linguistic markers or behavioral signs. These traces are often physical manifestations of behavior in the long term: for example, a little conscientious individual could often forget to water his plants, and a dried up plant in a room would be a behavioral trace. Of course, as some researchers recommend (Mehl, Gosling, Pennebaker, 2006), we must observe behaviors “in action”, that is to say during their course. These researchers propose that the observation of individuals in ordinary situations – their “train-train” ( humdrum lives) – helps us to better understand the link between personality and behavior. However, a major problem is that the collection of data in natural environment, as well as the codification of this content after the collection, are long and difficult tasks with traditional search tools (films, notes, audio recordings, photos … ). Tracking an individual closely and filming his / her behaviors is a laborious process which, therefore, greatly reduces the size of the samples. Recent technologies are somewhat alleviating these problems. For example, in their study of the link between personality and everyday language, the researchers mentioned previously (Mehl et al., 2006) used an automatic voice recorder, programmed to record a participant’s sound environment for 30 seconds every 12 minutes. Dictionary-based software was then used to produce quantitative linguistic data from these records.
Behavioral data collection in virtual worlds
9Virtual worlds offer unique advantages for studying the link between personality and behavior. For the purpose of this contribution, we define virtual worlds as visual digital environments that allow geographically separated individuals to interact through their avatars (ie, their digital bodies). It is important to mention that these virtual worlds have surpassed the academic prototypes and cultural niches of their origins: World of Warcraft has more than 11 million subscribers (White, 2009), and the Farmville gameon Facebook has more than 80 million active users. Virtual worlds have three characteristics that make them particularly suitable for collecting natural behavioral data. First, unlike the physical world where it would be impossible to follow everyone anywhere with a video camera, the virtual worlds are instrumented by nature: the computer systems that make the virtual world work already track the movement and the behavior of each avatar to allow interactions between the players (for example, to direct the avatars so that they look at each other). Then, these “virtual sensors” work permanently: this allows the collection not only of “slices of activity” but also of longitudinal samples, that can reveal important trends over time (Ducheneaut, Yee, Nickell, Moore, 2007). Finally, these observations take place in a transparent way, without visible impact on the users – which greatly reduces the effects of observation (Webb, Campbell, Schwartz, Sechrest, 1966): the participants can not change their behavior in front of the camera when the camera is invisible. Taking advantage of these features, a recent study illustrates the link between personality and virtual behavior in the world of participants can not change their behavior in front of the camera when the camera is invisible. Taking advantage of these features, a recent study illustrates the link between personality and virtual behavior in the world of participants can not change their behavior in front of the camera when the camera is invisible. Taking advantage of these features, a recent study illustrates the link between personality and virtual behavior in the world ofSecond Life (Yee, Harris, Jabon, Bailenson, in press). In this study, 76 students were recruited to participate in Second Lifeduring 6 weeks while “wearing” on their avatar a virtual object loaded with a script of capture of the motor and verbal behavior of their character. The results contain interesting correlations. For example, being conscientious is positively related to the amount of movement in the virtual world. However, this study has some limitations. First, it is difficult to capture truly natural data when participants are asked to use a virtual world that they may not be regular users of. Observing usual users would probably have led to more reliable results. Secondly, the data was collected in only one virtual world. If we consider that the vast majority of Second Lifelooks like an American suburb (Au, 2010), it might be useful to get more diverse environment data ( eg the fantasy world of an online game) to see if the results can be generalized. Thirdly, the participants in this study had to spend only 6 hours a week in Second Life – and we know that players from worlds like World of Warcraft spend much more time (20 hours / week), without external constraints. in other environments (Yee, 2006). In other words, these participants may not be representative. Finally, a large number of correlations present in this study are insufficiently aligned with the definitions of each personality trait – egthe virtual behaviors correlated with a pleasing personality do not seem to reflect the definition of this trait in the literature. As a result, a review of this study (or similar study) in a different virtual world and with regular users could clarify whether the results are an artifact of Second Life’snature or whether they reflect how users behave online in general.