Analysing Virtual Labs Through Integrated Multi-Channel Eye-Tracking Technology: A Proposal for an Explanatory Fit Model

Abstract

This study deals with an analysis of the cognitive load indicators produced in virtual simulation tasks through supervised and unsupervised machine learning techniques. The objectives were (1) to identify the most important cognitive load indicators through the use of supervised and unsupervised machine learning techniques; (2) to study which type of task presentation was most effective at reducing the task’s intrinsic load and increasing its germane load; and (3) to propose an explanatory model and find its fit indicators. We worked with a sample of 48 health sciences and biomedical engineering students from the University of Burgos (Spain). The results indicate that being able to see the task before performing it increases the germane load and decreases the intrinsic load. Similarly, allowing students a choice of presentation channel for the task respects how they process information. In addition, indicators of cognitive load were found to be grouped into components of position, speed, psychogalvanic response, and skin conductance. An explanatory model was proposed and obtained acceptable fit indicators.

Keywords

• eye tracking
• galvanic skin response
• cognitive load
• simulation tasks
• machine learning techniques