Eye-tracking data quality as affected by ethnicity and experimental design

Area of Science:

• Human-Computer Interaction
• Biomedical Engineering
• Computer Vision

Background:

• Accurate eye-tracking is crucial for reliable data collection and intended user actions.
• Suboptimal eye-tracking accuracy can lead to misinterpretations and incorrect system responses.
• Understanding factors influencing eye-tracking data quality is essential for diverse user populations.

Purpose of the Study:

• To evaluate trackability, accuracy, and precision of eye-tracking data across different ethnic groups.
• To investigate the impact of varying head positions and illumination conditions on eye-tracking performance.
• To identify key factors affecting eye-tracking data quality.

Main Methods:

• Collected eye-tracking data from participants of Asian, African, and Caucasian ethnicities.
• Measured trackability, accuracy, and precision under diverse head positions and lighting.
• Analyzed data quality indicators in relation to operating distance, illumination, and gaze angles.

Main Results:

• Asian participants exhibited lower accuracy and precision compared to African and Caucasian participants.
• Operating distance was the most significant factor affecting all data quality indicators across all ethnic groups.
• Illumination had no significant effect on accuracy or precision; dark backgrounds improved accuracy for African and Caucasian participants.
• Large gaze angles negatively impacted trackability for African participants and accuracy/precision for two ethnic groups.

Conclusions:

• Ethnic background influences eye-tracking accuracy and precision, necessitating tailored calibration or algorithms.
• Optimizing operating distance is critical for enhancing eye-tracking data quality across diverse populations.
• Gaze angle limitations should be considered, especially for specific ethnic groups and trackability requirements.