Table 3 - uploaded by Katrina Fong
Content may be subject to copyright.
Source publication
Many individuals now meet and develop friendships while online. As a result, people must form impressions of online acquaintances based on that person's online representation. Here, we investigate personality inferences and intentions to befriend based solely on simple avatars (i.e., customized cartoon representations of the self). Our data show th...
Similar publications
Citations
... While there has been significant work done on how people experience using avatars in virtual worlds, e.g., in terms of embodiment, presence, immersion, identification, self-identity, online friendships, and community (e.g., Badrinarayanan et al., 2014;Bülow & Felix, 2016;Crick, 2011;Ess, 2012;Farrow & Iacovides, 2014;Fong & Mar, 2015;Gies, 2008;Hardesty, 2016;Hilvoorde & Pot, 2016;Klevjer, 2012Klevjer, , 2022Nilsson et al., 2002;Salinäs, 2002;Schectman, 2012;Schroeder, 2002;Schultze, 2010;Tartaglia, 2012;Taylor, 2002), 1 there is comparatively little literature investigating our ability to express ourselves and understand each other through avatar interactions. Where the social implications of avatars are considered, it is commonly supposed that not only are our social encounters significantly altered when conducted through avatars but that the use of avatars hampers our ability to understand one another. ...
Critics have argued that human-controlled avatar interactions fail to facilitate the kinds of expressivity and social understanding afforded by our physical bodies. We identify three claims meant to justify the supposed expressive limits of avatar interactions compared to our physical interactions. First, “The Limited Expressivity Claim”: avatars have a more limited expressive range than our physical bodies. Second, “The Inputted Expressivity Claim”: any expressive avatarial behaviour must be deliberately inputted by the user. Third, “The Decoding Claim”: users must infer or figure out the expressive meaning of human-controlled avatars’ behaviour through cognitively onerous processes. With the aim of critically assessing all three claims, we analyze data collected through observations of and interviews with expert players of the avatar-based video game League of Legends. Focusing on Daniel Stern’s (2010) notion of vitality, we analyze the participants’ descriptions of seeing and interacting with other avatars during performance. Our analysis shows that the informants experience human-based avatarial interactions as qualitatively different than interactions with bots, that the informants see the movements of other players’ avatars as having different expressive styles, and that the informants actively use and manipulate this avatarial expressivity during performance. The results of our analysis, we argue, provide reasons for loosening or resisting the three claims concerning the limits of avatarial expressivity.
... When the virtual character depicts the user as self-presentation, for example, in computer games, it is called an avatar. 57 The term virtual agent refers to the dynamic visualization of a virtual character, and if connected to a conversational model, represents a conversational agent which is "an autonomous construct that maintains its own representations of the state of the world and the conversation, and whose behavior is determined by these representations". 58 A virtual being in our work differs from Cassell's et al. 59 definition of an embodied conversational agent (ECA) in that the latter represents a "computer-generated cartoon-like character" with a full body, whereas our virtual being conceptualizes a realistically "human-looking character," as targeted in research on virtual humans, 37,33,60 and only requires partial embodiment with the face. ...
The present work investigates the effect of natural conversations with virtual beings on user perceptions with a current conversational AI model (Meta's BlenderBot). To this aim, we designed a virtual being from a deep learning‐generated face and a conversational AI model acting as a virtual conversation partner in an online conferencing software and evaluated it in 11 perceptions of social attributes. Compared to prior expectations, participants perceived the virtual being as distinctly higher in warmth (engaging, empathic, and approachable) but lower in realism and credibility after 5 days of 10 min daily conversations (Study 1). Further, we explored the idea of simplifying the technical setup to reduce the technical entry barrier for such AI applications (Study 2). To this aim, we conducted several trials of fine‐tuning a small conversational model of 90 million parameters until its performance metrics improved. Testing this fine‐tuned model with users revealed that this model was not perceived differently from a large conversational model (1.4 billion parameters). In summary, our findings show that recent progress in conversational AI has added warmth‐related aspects to the user experience with virtual beings, and that fine‐tuning a conversational AI model can be effective to reduce technical complexity. We created a virtual being composed of a conversational AI model (run on a cloud‐based GPU) that was visualized by a talking and listening AI‐generated face. We let participants talk to this virtual being in Zoom for five days. Results revealed that Participants evaluated most perceptions of the virtual being as lower than their expectations. Warmth‐related perceptions (approachable, engaging, empathetic) were rated higher than expectations. These three perceptions formed a separate psychological dimension.
... This certainly includes the social media phenomenon of catfishing, or when a person creates a fictional persona or fake identity to target and deceive specific victims (Simmons & Lee, 2020). We also suspect that this line of research can inform the tangential topics of self-validation theory, i.e., the cognitive and affective validation of one's thoughts (see Briñol & Petty, 2021) and avatars, or the virtual representations of selves in artificial and mediated environments (Fong & Mar, 2015;Sah et al., 2021;Wu, 2013). People indeed tend to behave over time in ways consistent with their avatars' characteristics, a phenomenon known as the Proteus effect (Ratan et al., 2020). ...
‘Brushy Bill’ Roberts gained notoriety in 1950 for identifying himself as the presumably deceased outlaw ‘Billy the Kid.’ We hypothesized that his case reflected extreme celebrity worship, which involves psychological absorption with a target celebrity and potential fantasy-reality breakdowns. A blinded expert panel mapped Roberts’ claims, activities, and circumstances against the three phases of celebrity worship and their known correlates. Outcomes from this exercise suggested that: (a) his reported attitudes and behaviors equated to an above-average score on the Celebrity Worship Scale (McCutcheon, Lange, & Houran, 2002), (b) his identity as the Kid unfolded somewhat similarly to the behavioral progression of celebrity worship, and (c) he ostensibly had the most psychosocial risk factors for the ‘Entertainment-Social’ level of celebrity worship, though many were also noted for the more extreme ‘Intense-Personal’ and ‘Borderline Pathological’ phases. These results imply that Roberts might have consciously adopted Billy the Kid as an alter ego primarily for leisure and escapism, although this construction perhaps evolved to include more compulsory or addictive aspects.
Keywords: absorption, alter ego, celebrity worship, grandiose delusions, narrative reality
... The physical appearance of avatars can be simple or complex: users can indeed represent themselves through a very simple drawing (e.g., Mii characters for the Nintendo Wii) or three-dimensional figures rich with details (e.g., video games) (Fong & Mar, 2015). It is important to emphasize that avatars are not exclusive to virtual worlds and computer games, rather they exist (e.g., in their "relational" version) in old and basic social media as well, such as forums and webchats. ...
... For example, introverted people and neurotic women tend to select more attractive avatars, while participants with low self-esteem picked avatars with lighter skin tones (Dunn & Guadagno, 2012). One study demonstrated that participants were able to guess other participants' personality traits just by observing their avatars (Fong & Mar, 2015) and this shows that the connection between players and avatars is strong and meaningful; however, there are situations in which avatars are very different from their users . For example, the research showed that users may (or may be primed to) create avatars resembling an ideal representation of themselves (Jin, 20092010), and that a noticeable discrepancy between an ideal and an actual self-virtual representation could be a sign of psychological and health issues (Kim & Sundar, 2012;Leménager et al., 2013). ...
... As mentioned earlier, Fong and Mar (2015) conducted an experimental study to investigate whether people generally try to create avatars that represent themselves accurately, or whether they aim to display themselves differently than they appear in real life. Fifty subjects were asked to create an avatar, while the other 50 were specifically asked to create an avatar that would represent their personality accurately. ...
Avatars are an important feature of digital environments. Existing both in social networks and webchats (usually as static images) and in single-player and online video games (as dynamic characters, often humanoid), avatars are meant to represent users' action and communication within digital environments. Research has shown that, when they are customized by users, avatars are not created “randomly,” rather they maintain some kind of relationship with users' actual self-representation and identity. However, more recent studies showed that users may have multiple digital representations: the same person could create multiple avatars depending on which facet of the self is primed by an experimental manipulation, or on which aims they have to pursue in the given virtual environments (e.g., to seduce, to play, to work). With this background, this contribution explores the possibility to use customized avatars within psychological assessment, as adjunctive assessment tools useful to get information on patients' self-representation(s) and communicative intentions.
... A myriad of research has explored how people use avatars to shape their self-images in virtual social spaces, including both PC-based virtual worlds [41] and the more immersive social VR [25,26]. Research has found that while many people designed their avatars to reflect their physical appearance in the real world [57], some people leveraged the opportunity to experiment with different aspects of their personalities [18,24] and even explored completely different identities [81]. However, most prior works focused on the self-presentation preferences of people without disabilities. ...
In social Virtual Reality (VR), users are embodied in avatars and interact with other users in a face-to-face manner using avatars as the medium. With the advent of social VR, people with disabilities (PWD) have shown an increasing presence on this new social media. With their unique disability identity, it is not clear how PWD perceive their avatars and whether and how they prefer to disclose their disability when presenting themselves in social VR. We fill this gap by exploring PWD's avatar perception and disability disclosure preferences in social VR. Our study involved two steps. We first conducted a systematic review of fifteen popular social VR applications to evaluate their avatar diversity and accessibility support. We then conducted an in-depth interview study with 19 participants who had different disabilities to understand their avatar experiences. Our research revealed a number of disability disclosure preferences and strategies adopted by PWD (e.g., reflect selective disabilities, present a capable self). We also identified several challenges faced by PWD during their avatar customization process. We discuss the design implications to promote avatar accessibility and diversity for future social VR platforms.
... This accuracy has been explained by individuals' personality traits being reflected in their verbal and nonverbal behavior, facial expressions, or looks (Gosling et al., 2002;Naumann et al., 2009). For example, agreeable people are more likely to have a friendly/smiling profile picture on social media (Hall et al., 2014) and to create avatars that others would want to befriend (Fong & Mar, 2015); in their everyday conversations, agreeable people are less likely to swear and are more likely to express positive emotions (Mehl et al., 2006;Yarkoni, 2010). Hence, spending 20 min working together could have allowed individuals to pick up on such cues and use them to make trust decisions. ...
Existing research has documented the social benefits (i.e., higher popularity and liking) of extraversion and agreeableness. Do these positive reputational consequences extend to social dilemma situations that require trust? We found that people do not trust extraverts more than introverts. Instead, people's trust decisions are guided by their partner's level of agreeableness. In a trust game (Studies 1 and 2), individuals were more likely to trust a partner who was described as agreeable (vs. disagreeable); and, in a laboratory study of work groups, participants trusted more (vs. less) agreeable group members (Study 3). Individuals anticipated others' preferences for agreeable partners and tried to come across as more agreeable, but not more extraverted, in social dilemmas (Study 4). These findings suggest that the social benefits of agreeableness (but not extraversion) extend to social interactions involving trust and highlight the importance of target personality traits in shaping trust decisions.
... It possibly signified that GPR was a relatively suitable algorithm for DOTA 2 behavioral datasets. Some researchers have investigated how to predict players' psychological traits through in-game behavior, but the results did not show strong predictive power (Yee et al., 2011a;Fong and Mar, 2015;Bunian et al., 2017;Wu et al., 2021). For example, Bunian et al. (2017) used hidden Markov models (HMM) to extract the sequence of players' actions and trained classification models for the Big Five personality traits. ...
As traditional methods such as questionnaires for measuring risk propensity are not applicable in some scenarios, a nonintrusive method that could automatically identify individuals' risk propensity could be valuable. This study utilized Defense of the Ancients 2 (DOTA 2) single match data and historical statistics to train predictive models to identify risk propensity by machine learning methods. Self-reported risk propensity scores from 218 DOTA 2 players were paired with their behavioral metrics. The best-performing model occurred with Gaussian process regression. The root mean square error of this model was 1.10, the correlation between predicted scores and self-reported questionnaire scores was 0.44, the R-squared was 0.17, and the test–retest reliability was 0.67. We discussed how selected behavioral features could contribute to predicting risk propensity and how the approach could be of potential value in the application of perceiving individuals' risk propensities. Moreover, the limitations of our study were discussed, and recommendations were made for future studies in this field.
... Fong et al. [55] Discussed personality inferences and intentions to befriend based solely on online avatars ...
With the emergence of personality computing as a new research field related to artificial intelligence and personality psychology, we have witnessed an unprecedented proliferation of personality-aware recommendation systems. Unlike conventional recommendation systems, these new systems solve traditional problems such as the cold start and data sparsity problems. This survey aims to study and systematically classify personality-aware recommendation systems. To the best of our knowledge, this survey is the first that focuses on personality-aware recommendation systems. We explore the different design choices of personality-aware recommendation systems, by comparing their personality modeling methods, as well as their recommendation techniques. Furthermore, we present the commonly used datasets and point out some of the challenges of personality-aware recommendation systems.
... As a consequence, different emojis may also exert varying effects on interpersonal trust. According to the research on computermediated communication, cues from online virtual avatars can provide an accurate perception of extroversion and agreeableness of personality traits (Kim et al., 2012;Fong and Mar, 2015;Aseeri and Interrante, 2021). Compared with other clues, virtual images without expression can negatively influence initial online trust among investors. ...
Emojis are increasingly used in online communication and expression, however, most previous studies have focused on describing this phenomenon, but less on how it affects interpersonal trust relationships. Therefore, this study examines the effect of emojis on online interpersonal trust among college students through three experiments. A total of 62 college students were recruited for Experiment 1. The results demonstrated that positive emoji () improved the level of trust of trustors in the trust game [t(60) = -2.79, p = 0.007], whereas that of the control group exerted no effect on the initial level of online trust among college students. Then, 74 college students were selected for Experiment 2. The results indicated no significant differences between the experiment and control groups in terms of the influence of negative emojis () on initial online trust using. A joint analysis (via ANOVA) of Experiments 1 and 2 illustrated that the type of emoji exerted a significant effect [F(2,96) = 3.96, p = 0.02, η 2 = 0.08] on college students' online trust. Finally, we recruited 111 participants for Experiment 3 to explore the role of emojis on online trust among acquaintances. The results suggested that the individual propensity to trust plays a moderate role in the relationship between emojis and online trust among acquaintances. That is, emojis influenced interpersonal trust among acquaintances only if the level of propensity to trust, is low.
... For example, in a study using avatars created for Second Life, Bélisle and Bodur (2010) found that people correctly identify specific avatar cues related to the traits of extraversion and agreeableness (e.g., avatars with stylish hair were correctly perceived to be created by more extraverted users; and attractive and feminine looking avatars were correctly perceived to belong to more agreeable users). In addition, Fong and Mar (2015) found that people were able to judge extraversion, agreeableness, and neuroticism from pixelated avatars from weeworld.com at a better than chance level (with accuracy correlations from r = 0.10 to 0.24). ...
... Sensitivity analyses showed we had at least 80% power to detect a small accuracy effect, a correlation between perceived and actual trustworthiness of r = 0.15 or greater, suggesting that power might not be an issue. In addition, our sample of avatars (N = 360) was substantially larger than samples used in previous studies examining the accuracy of personality impressions in avatars (N's = 99 and 103 avatars: Bélisle & Bodur, 2010;Fong & Mar 2015). Our sample of raters (55-60 raters per avatar) was also considerably larger than needed to obtain reliable measures of perceived trustworthiness (i.e., 20-30 raters per stimulus; Hehman, Xie, Ofosu, & Nespoli, 2018). ...
In some online interactions, people use avatars to represent themselves and judge whether interaction partners should be trusted. However, little is known about human accuracy in perceptions of avatar trustworthiness. We conducted a two-stage study to investigate whether people are able to accurately judge trustworthiness in avatars. In Stage 1, participants created avatars using avatarmaker.com and made decisions as trustees in an incentivized trust game (N = 360 avatars). In Stage 2, a new group of participants (N = 315 raters) were presented with avatars created in Stage 1; rated their trustworthiness; and made decisions about whether to trust them. We observed three main results: First, there was a high level of consensus in perceptions of avatar trustworthiness. Differences in trustworthiness judgments were mainly attributed to differences between avatars (rather than differences between raters). Second, raters' trust decisions were guided by their perceptions of avatar trustworthiness. Third, perceptions of avatar trustworthiness were not associated with actual trustworthiness of avatar creators. People were not, on average, able to accurately identify the actual trustworthiness of avatar creators from avatar appearance. Our results suggest that people erroneously rely on others’ avatar appearance in online interactions.
