js-mEye: An extension and plugin for the measurement of pupil size in the online platform jsPsych

The measurement of pupil size has become a topic of interest in psychology research over the past two decades due to its sensitivity to psychological processes such as arousal or cognitive load. However, pupil measurements have been limited by the necessity to conduct experiments in laboratory settings using high-quality and costly equipment. The current article describes the development and use of a jsPsych plugin and extension that incorporates an existing software that estimates pupil size using consumer-grade hardware, such as a webcam. We validated this new program (js-mEye) across two separate studies, which each manipulated screen luminance and color using a novel luminance task, as well as different levels of cognitive load using the N-back and the Stroop tasks. Changes in luminance and color produced significant changes in pupil size in the hypothesized direction. Changes in cognitive load induced in the N-back and Stroop tasks produced less clear findings; however, these findings were explained to some extent when participant engagement - indexed by task performance - was controlled for. Most importantly, all data were at least moderately correlated with data simultaneously recorded using an EyeLink 1000, suggesting that mEye was able to effectively substitute for a gold-standard eye-tracking device. This work presents an exciting future direction for pupillometry and, with further validation, may present a platform for measuring pupil size in online research studies, as well as in laboratory-based experiments that require minimal equipment.