Development of a Novel Transition Frequency Eigenvalue/Principal Component Approach for the Analysis of Eye-Tracking Data
Dr. Justin Shorb
The advent of digital multimedia resources in education has required careful thought as to the best methods for organizing and creating them so as best to help students learn complex concepts. This issue is highlighted within the field of Chemistry Education reflecting the need to relay concepts that cannot be visualized without some level of abstraction and representation. Recent work in understanding student engagement with digital tools has led to the use of eye-tracking technology to monitor student gaze patterns. Until recently, it has been impossible to correlate gaze patterns across more than two areas of interest (a paragraph, an equation, and a picture would be three areas of interest). Our group’s novel transition-frequency principle component analysis method allows for more complex coupled gaze patterns to be quantified. This method uses eigendecomposition of transition frequency matrices to find correlations in viewing patterns. In collaboration with our collaborators at GVSU, initial benchmarking and consistency testing of the new method will be demonstrated.
A recommended citation will become available once a downloadable file has been added to this entry.