How comprehensible are eye-movement modelling examples? What annotations can tell

Abstract

Eye movement modelling examples (EMMEs) are a type of video-based instruction that uses a model’s gaze to draw the viewer’s attention directly to the procedures that are being taught. Its uses extend across areas such as sports, aviation, medical education and reading instruction. The assumption is that watching a model may prompt the viewer to apply these procedures in similar tasks in the future, and visualizing good and poor behavior can trigger contrasting analysis. While the effects of EMMEs are usually measured in transfer tasks and model-observer similarity, few studies employed online comprehension checks. In the present study, we analyze the annotation protocol developed by do Amaral et al. (2025), who used EMMEs to foster digital reading competencies. The notes consisted of model descriptions and justified ratings. We hypothesized that annotations were not only a comprehension check but also a metacognitive tool that would foster processing of the EMMEs. We also investigated a possible correlation between annotation scores and performance in the transfer tasks of the previous study. Results indicated that participants understood the video models, as evidenced by high scores in the descriptions and ratings. However, qualitative analyses of the justifications revealed that participants attended to content features that were not fixated by the model. Content salience is justified in light of the purpose of the EMMEs, which aimed at developing digital reading competencies. Annotation scores did not correlate with transfer measures. We interpret this result as indicative of the discreteness between metacognition and observational learning, with the effect of EMMEs standing independently from written elaboration processes.