To Be or Not to Be Human-Like in Virtual World - Aix-Marseille Université Accéder directement au contenu
Article Dans Une Revue Frontiers in Computer Science Année : 2020

To Be or Not to Be Human-Like in Virtual World

Résumé

The main objective is a double one. First and foremost, it is a question of showing that foot-in-the-door as a proven behavioral influence technique in offline interactions maintains its efficiency in online interactions. It is then a question of exploring the impact of the anthropomorphism vs. the non-anthropomorphism of the requester avatar on the efficiency of this technique. Foot-in-the-door is based on a simple principle: you start by asking for a little in a first step to increase the probability of obtaining a lot in a second step. The research was conducted in the Second Life virtual world. In a control condition (n = 200), a requester avatar directly proposed the target request. In a foot-in-the-door condition (n = 200), the requester avatar started by presenting a preparatory request before proposing the target request. According to the conditions, the requester avatar was human-like (female or male), or non-human-like (flower, balloon, cube). As expected, our results show that overall the foot-in-the door-technique remains efficient in the virtual world; they also show that this efficiency depends on the human-like form of the requester. avatar. This last result is interpreted as a reference to the theory of social presence. Non-human-like avatars could generate a weak social presence, to the point where the mechanisms of self-perception and commitment underlying the foot-in-the-door effect may not be automatically initiated. Player avatars would in this way be freed from the rules of social interaction occurring in offline interactions

Dates et versions

hal-02988067 , version 1 (04-11-2020)

Identifiants

Citer

Laura Barbier, Valérie Fointiat. To Be or Not to Be Human-Like in Virtual World. Frontiers in Computer Science, 2020, 2, ⟨10.3389/fcomp.2020.00015⟩. ⟨hal-02988067⟩
56 Consultations
0 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More