An Experimental Study on the Virtual Representation of Children (Presented as Poster in IEEE VR 2015)

Is it their movements or appearance that helps us to identify a child as a child? We created four video clips with a Virtual Character walking, but with different combinations of either child or adult animation applied on either a child or adult body. An experimental study was conducted with 53 participants who viewed all four videos in random orders. They also reported higher level of empathy, care, and feeling of protection towards the child character as compared to the adult character. Moreover, compared to appearance, animation seems to be playing a bigger role in invoking participants’ emotional responses.

A link to a webpage containing the 4 video clips and questions were sent and 53 participants who each gave rated all 4 videos.

• Is this a Child: 96% participants identify the CaCb as a child. A repeated measures two-way ANOVA test shows that both Animation and Body (p<0.00001) have an effect on participants’ judgments.
• Age group estimation: A repeated measures two-way ANOVA showed that both appearance and animation have a significant effect on age estimation (p<0.00001) with an interaction effect (p<0.035): when Child Body and Child Animation are together, the effect is more than the separate effect of each of them.
• Empathy, care, and feeing of protection: the CaCb video triggered the highest of all scores, with AaAb video the lowest. A repeated measures two-way ANOVA test shows that Animation has an effect on the level of empathy (p<0.0001). Body also has an effect on empathy but with a lower significant level (p<0.05).
• With the consistent videos, the child character triggered significant higher level of sympathy in the participants as compared to the adult character When there is a conflict between behavior and appearance, participants’ judgments and emotional response were more likely to be dominated by the animation behavior rather than appearance.