Using Identification with AR Face Filters to Predict Explicit & Implicit Gender Bias

Abstract

Augmented Reality (AR) filters, such as those used by social media platforms like Snapchat and Instagram, are perhaps the most commonly used AR technology. As with fully immersive Virtual Reality (VR) systems, individuals can use AR to embody different people. This experience in VR has been able to influence real world biases such as sexism. However, there is little to no comparative research on AR embodiment’s impact on societal biases. This study aims to set groundwork by examining possible connections between using gender changing Snapchat AR face filters and a person’s predicted implicit and explicit gender biases. We discovered that participants who experienced identification with gender manipulated versions of themselves showed both greater and lesser amounts of bias against men and women. These results depended the user’s gender, the filter applied, and the level of identification users reported with their AR manipulated selves. The results were similar to past VR findings but offered unique AR observations that could be useful for future bias intervention efforts.

Marie Jarrell

Postdoc

Hi, I’m Marie, a researcher, designer, and staunch advocate for all things related to video games, VR simulations, and overall digital interactive experiences. A graduate of Clemson University with a PhD in Human-Centered Computing and two Master degrees in Digital Production Arts and Computer Science. I currently utilize my numerous artistic and scientific skills to build video games and research their impact.

Etienne Peillard

Associate Professor

My research interests include human perception issues in Virtual and Augmented Reality, spatial perception in virtual and augmented environments, and more generally, the effect of perceptual biases in mixed environments.