Investigation of Web-Based Eye-Tracking System Performance under Different Lighting Conditions for Neuromarketing

Abstract

The increasing popularity of neuromarketing has led to the emergence of various measurement methods, such as webcam-based eye-tracking technology. Webcam-based eye-tracking technology is noteworthy not only for its use in laboratories but also for its ability to be applied to participants online in their natural environments through a link. However, the complexity of e-commerce interfaces necessitates high performance in eye-tracking methods. This complexity and the applicability of webcam-based eye-tracking technology in various environments have raised research questions about how its performance changes depending on the type and location of lighting. To answer these questions, experiments were conducted with 30 users in two different experimental environments illuminated by artificial and natural methods, with the lighting from the left, right, and front. Participants were asked to focus on targets located in specially prepared graphics for the experiment. In the heatmaps obtained in the eye-tracking tests, the distance and angular difference between the focal point and the target point were measured using the polar coordinate system. The findings indicate that measurements taken with lighting coming from the center were more efficient in both natural and artificial lighting types and measurements taken under natural lighting performed 24% better than artificial ones. Web camera-based eye-tracking technology is a promising method. However, detailed statistical analyses have demonstrated that for complex interfaces like e-commerce, the position and type of lighting are crucial parameters.