Consequences of Eye Color, Positioning, and Head Movement for Eye-Tracking Data Quality in Infant Research

Abstract

Eye tracking has become a valuable tool for investigating infant looking behavior over the last decades. However, where eye-tracking methodology and achieving high data quality have received a much attention for adult participants, it is unclear how these results generalize to infant research. This is particularly important as infants behave different from adults in front of the eye tracker. In this study, we investigated whether eye physiology, positioning, and infant behavior affect measures of eye-tracking data quality: accuracy, precision, and data loss. We report that accuracy and precision are lower, and more data loss occurs for infants with bluish eye color compared to infants with brownish eye color. Moreover, accuracy was lower for infants positioned in a high chair or in the parents' lap compared to infants positioned in a baby seat. Finally, precision decreased and data loss increased as a function of time. We highlight the importance of data quality when comparing multiple groups, as differences in data quality can affect eye-tracking measures. In addition, we investigate how two different measures to quantify infant movement influence eye-tracker data quality. These findings might help researchers with data collection and help manufacturers develop better eye-tracking systems for infants. Copyright