Eye movements in the manipulation of hands-on and computer-simulated scientific experiments: an examination of learning processes using entropy and lag sequential analyses

Computer-simulated experiments have been gaining popularity over hands-on experiments in science education, given the availability of technology and the trend of distance learning. Past studies have focused primarily on comparing the learning outcomes and user experiences of the two experiment modes. In this study, we used an eye tracker to investigate the learning processes involved in manipulating hands-on and computer-simulated experiments, and the effect of prior knowledge and experiment mode on eye movements. A total of 105 undergraduates completed either mode of experiment to learn about pulley mechanics. Participants were asked to read relevant concepts before conducting the experiments to ensure they had basic knowledge about the subject matter. Results showed that the learning outcome of experimentation was affected by prior knowledge but not experiment mode. As for eye movements, the two experiment workstations were divided into nine functional regions. The findings revealed that eye movements in most regions were affected by the experiment mode, but not prior knowledge. The simulation group had shorter total fixation durations and smaller pupil sizes than the hands-on group, implying a lower cognitive load in learning in computer-simulated experiments. Lag sequential analysis and entropy analysis were conducted on cross-regional fixation transitions. The results revealed that participants in hands-on experiments tended to make more diversified fixation transitions across regions, whereas those in simulated experiments showed a higher level of concentration in the spatial pattern of fixation transitions. While sequential analysis offers insights into important fixation transitions on a regional level, entropy analysis allows for a more macro perspective on the overall transition distribution and facilitates conventional statistical modeling that takes individual differences into account.