The influence of calibration method and eye physiology on eyetracking data quality

Area of Science:

• Oculomotor research
• Human-computer interaction
• Biomedical engineering

Background:

• High-quality eye movement data is essential for valid and replicable research findings.
• Current data quality assessment in eye-tracking is often informal and lacks quantification.
• Understanding factors influencing eye-tracking data quality is critical for advancing oculomotor system research.

Purpose of the Study:

• To systematically investigate factors affecting eye-tracking data quality.
• To quantify the impact of calibration methods, participant physiology, and recording conditions on data accuracy, precision, and validity.
• To provide empirical evidence for improving eye-tracking data acquisition protocols.

Main Methods:

• Systematic investigation of a tower-mounted, video-based eye-tracker.
• Quantification of data quality metrics including accuracy, precision, and percentage of valid data.
• Analysis of influences from calibration method, participant eye physiology (glasses, contact lenses, eye color, eyelashes, mascara), recording time, and gaze direction.

Main Results:

• Participant-indicated calibration significantly improved data quality compared to operator or software-guided calibration.
• Factors such as glasses, contact lenses, eye color, eyelashes, and mascara were found to statistically influence data quality.
• The study identified key variables impacting the accuracy, precision, and validity of recorded eye movement data.

Conclusions:

• Participant-led calibration is recommended for optimizing eye-tracking data quality.
• Awareness of physiological factors is necessary for researchers to mitigate potential biases in eye-tracking data.
• The findings offer practical guidance for eye movement researchers and eye-tracker manufacturers to enhance data quality and device performance.