Abstract
Accurate diagnosis of student misconceptions is crucial for effective science instruction. Traditional diagnostic tools, however, often rely on unreliable self-reported confidence levels. This study developed and validated a two-dimensional diagnostic model that integrates objective eye tracking data with test responses to more robustly identify misconceptions in Newtonian mechanics. The study involved three phases: developing a four-tier diagnostic test, conducting an eye tracking experiment to analyze visual attention patterns (fixation duration, revisits, scan paths) among high school students, and constructing the final model by correlating eye movements with confidence judgments. Consistent with previous research, results showed that students with misconceptions had shorter fixation durations on correct options and displayed inefficient scan paths. Conversely, those with scientific conceptions showed focused attention. As a novel contribution of this study, a “reversed U-shaped” pattern was observed between confidence and revisits, indicating that revisits can serve as an objective measure of cognitive uncertainty. Based on these findings, we propose a two-dimensional “eye tracking-test” model. This model offers a more precise method for diagnosing student understanding, providing a foundation for targeted instructional interventions and fostering a dynamic “teach-learn-assess” feedback loop for personalized physics education.
Funding Information
  • National Natural Science Foundation of China (U23A2075)
  • National Natural Science Foundation of China (12274132)
  • Fundamental Research Funds for the Central Universities
  • Shanghai Municipal Education Commission (202101070008E00099)
  • 2023 Association of Higher Education of P.R. China (23XXK0202)