Online webcam-based eye tracking in cognitive science: A first look.

Area of Science:

• Cognitive Psychology
• Human-Computer Interaction
• Web-based Research

Background:

• Online experimentation is increasingly used in cognitive psychology.
• Eye tracking, a key cognitive psychology tool, has been underutilized in online studies.
• Previous research focused on performance and reaction time in online experiments.

Purpose of the Study:

• To evaluate the feasibility of web technology-based eye tracking for cognitive psychology research.
• To compare the spatial precision and pattern replicability of online eye tracking with in-lab data.
• To assess the suitability of JavaScript-based eye tracking algorithms with consumer webcams.

Main Methods:

• Comparison of three tasks (fixation, pursuit, free viewing) conducted both in-lab and online.
• Utilized JavaScript-based eye tracking algorithms and consumer-grade webcams for online data collection.
• Analyzed spatial precision for fixation and pursuit tasks, and gazing pattern replicability for free viewing.

Main Results:

• In-lab fixation task data showed a 172 px offset; online data had a 207 px offset and higher variance.
• Pursuit task data revealed similar offsets for both in-lab (211 px) and online (216 px) conditions.
• Free viewing task successfully replicated known attention patterns, with eyes (28.25%) attracting more gaze than nose (9.71%) or mouth (4.00%).

Conclusions:

• Web technology-based eye tracking is a viable method for fixation, pursuit, and free viewing tasks in cognitive psychology.
• Online eye tracking offers a promising avenue for remote data collection, complementing traditional in-lab studies.
• Continued advancements in hardware and software will likely enhance the sophistication of web-based eye tracking paradigms.