Eye-gaze data to measure students’ attention to and comprehension of computational thinking concepts

Abstract

There is no agreed-upon assessment to evaluate learning of computational thinking concepts (CTCs), and none of the previous ones analyzed students’ eye-gaze measures over CTCs. This study investigates use of eye-gaze data over various CTCs, and tests the validity of eye-gaze data as an effective assessment instrument to measure learning of CTCs. Students’ eye-gaze data (FD: Fixation Duration, SA: Saccade Amplitude, RF: Regression Frequency) over CTCs were collected and correlated over seven sessions. The eye-gaze data were also correlated with quiz scores to test its validity in evaluating learning of CTCs. There was a statistically significant correlation between RFs on conditionals CTC and sessions showing increased frequency of regression in attention to conditionals, and between FDs and quizzes indicating FDs’ validity in measuring the learning of CTCs. These results implicate that FDs are effective in assessing the learning of CTCs and students are revisiting conditionals more often as compared to other CTCs.