Prediction of food choice based on eyemovement data

Abstract

Nowadays, we are exposed to many stimuli in all areas of our lives, all of which affect our decisions, including food choices. During decision-making, eye movement provides a lot of information, with the help of which the decisionmaking process can be understood and accordingly, for example, manufacturers can design the ideal packaging or the shelf image in retail units can be properly designed. The choice of product is a decisive step during the purchase, so its examination is important and informative.

During our measurement, we followed the eye movements of 61 participants using a static eye camera the participants were shown a series of 4 alternative forced choice (4AFC) pictures, the pictures contained four Milka chocolate bars with different flavors. The task of the participants was to choose the product they liked the most. During product selection, we measured the following parameters: number of fixations and visits, as well as their length, the time elapsed until the first fixation and the length of the first fixation. During the evaluation of the results, we established that the product alternative has a significant effect on eye movement parameters. The participants did not significantly look more at the selected product, that is, one participant did not look more at the product he selected than the other participant at the product he selected. Participants look significantly more at the product they have selected.

Based on the results, it can be concluded that the product samples did not influence the eye movement parameters except for the time elapsed until the first fixation. This exception can be explained by the location of the images, because most people start from the upper left corner when reading, and then proceed clockwise in the same direction. Based on the results, it can also be said that the chosen product really received more visual attention, both in the case of visits and fixations, but the time elapsed until the first fixation and the length of the first fixation did not influence the decision. In other words, the participants did not always choose the product they saw first. Examining the interactions between the eye movement parameters and the choices, it can be concluded that the raspberry-flavored Milka chocolate received the most visual attention, even during the very first visit. Strawberry-flavored Milka chocolate received the least visual attention, both in terms of visits and fixations.