Seeing Bias Through Eye Movements: Investigating Empathy and Engagement with Authentic and Altered Refugee Imagery
Publication Date
1-1-2025
Document Type
Article
Publication Title
IEEE Transactions on Human Machine Systems
DOI
10.1109/THMS.2025.3585452
Abstract
This study investigates perceptual biases in responses to Syrian and Ukrainian refugee images using human-computer interaction (HCI) techniques. We hypothesized that media portrayals might shape empathy and influence both biases and donation behavior. Participants viewed both authentic and face-swapped images, with eye-tracking data capturing pupil size and decision-making patterns. In Experiment 1, participants' eye-tracking data and donation behavior indicated higher engagement and donations for authentic images, suggesting a preference for unaltered images without ethnicity-based bias. Experiment 2 explored this further by presenting cropped refugee faces alongside neutral or intellectual art, asking participants to choose the art that best matched each face. Results showed no ethnic bias. Findings suggest that face-swapping reduces emotional engagement, and that facial authenticity is critical to empathy (Experiment 1), while no bias toward specific ethnic features emerged in the cropped face task (Experiment 2). This research underscores the value of HCI methods in uncovering implicit biases and supports the use of authentic imagery in media to foster empathy in humanitarian efforts.
Keywords
Bias, donation, eye-tracking, face swapping, ordinary procrustes analysis (OPA), pupil size, refugees, response time (RT)
Department
Computer Science
Recommended Citation
Reem Albaghli, Akshay Sunil Gurnaney, and Nada Attar. "Seeing Bias Through Eye Movements: Investigating Empathy and Engagement with Authentic and Altered Refugee Imagery" IEEE Transactions on Human Machine Systems (2025). https://doi.org/10.1109/THMS.2025.3585452
