Validation of Gazepoint low-cost eye-tracking and psychophysiology bundle

Hélio Clemente Cuve,Jelka Stojanov,Xavier Roberts-Gaal,Caroline Catmur,Geoffrey Bird

DOI: https://doi.org/10.3758/s13428-021-01654-x

Abstract:Eye-tracking and recording of physiological signals are increasingly used in research within cognitive science and human-computer interaction. For example, gaze position and measures of autonomic arousal, including pupil dilation, skin conductance (SC), and heart rate (HR), provide an indicator of cognitive and physiological processes. The growing popularity of these techniques is partially driven by the emergence of low-cost recording equipment and the proliferation of open-source software for data collection and analysis of such signals. However, the use of new technology requires investigation of its reliability and validation with respect to real-world usage and against established technologies. Accordingly, in two experiments (total N = 69), we assessed the Gazepoint GP3-HD eye-tracker and Gazepoint Biometrics (GPB) system from Gazepoint. We show that the accuracy, precision, and robustness of the eye-tracker are comparable to competing systems. While fixation and saccade events can be reliably extracted, the study of saccade kinematics is affected by the low sampling rate. The GP3-HD is also able to capture psychological effects on pupil dilation in addition to the well-defined pupillary light reflex. Finally, moderate-to-strong correlations between physiological recordings and derived metrics of SC and HR between the GPB and the well-established BIOPAC MP160 support its validity. However, low amplitude of the SC signal obtained from the GPB may reduce sensitivity when separating phasic and tonic components. Similarly, data loss in pulse monitoring may pose difficulties for certain HR variability analyses.