Visual Attention Patterns in Processing Compiler Error Messages

• Main Authors: Tablatin, Christine Lourrine, Rodrigo, Ma. Mercedes T.
• Format: text
• Published: Archīum Ateneo 2023
• Subjects: compiler error messages; debugging; eye-tracking; fixation metrics; visual attention; Computer Engineering; Engineering
• Online Access: https://archium.ateneo.edu/discs-faculty-pubs/401
• Institution: Ateneo De Manila University

The difference in the visual attention of subjects while performing a debugging task can be measured using fixation count and fixation duration metrics. Thus, this paper investigated the visual attention patterns of high and low performing students engaged in a defect-finding task on multiple programs using these metrics. We performed statistical tests on the proportional fixation durations and fixation counts on the error lines and the compiler error messages to determine the difference in the visual attention patterns between the groups. The results of the statistical analysis revealed a significant difference between the high and low performing students across all programs. This implies that high performing students were associated with significantly higher visual attention on the error lines of the programs than the low performing students. However, the analysis of the proportional fixation duration and fixation count on the compiler error messages revealed no significant difference between the groups. The results suggest that both groups showed similar visual attention to the compiler error messages. The findings of this study provide insights into the visual attention patterns of student programmers in processing compiler error messages. High and low performing students could be distinguished based on their visual attention patterns on the error lines but not on the compiler error messages. Further, high performing students prefer a more analytical processing approach and pay attention to relevant code elements in debugging to correctly identify the errors while low performing students choose a more holistic approach.