Understanding Graph Reading and Comprehension through Eye-Tracking: Evidence for the Expert/Novice Dichotomy

Monday 11:30am-1:30pm UMC Aspen Rooms
Poster Presentation Part of Geoscience Education Research


James Lindgren, Macalester College
Karl Wirth, Macalester College
Graphical information is used in many aspects of our lives, including vocation, media, civic processes, scientific inquiry, and education, so graph comprehension is an essential skill for informed citizenry. However, relatively little is understood about how individuals perform graph-reading tasks or how these skills develop over time. Furthermore, many different forms of graphical information are used in earth science courses (e.g., "upside down" binary plots with a depth variable increasing in a downward direction; log-scales; normalized trace element diagrams; ternary plots) and can present significant thresholds for student learning. Here, we describe the results from an on-going two-year collaborative study on the skills and challenges behind graph reading and scientific literacy. Our data provides interesting insights into the differences between and within expert and novice populations that we hope will eventually illuminate new ways for improving students' graph comprehension skills.
Within our expert and novice pools (distinguished by level of education), measures of the accuracy of graph interpretation show a clear dichotomy between the two groups. Experts (faculty and staff) are more accurate in interpreting graphical media. In comparison, novices (undergraduate students), regardless of their level of degree completion, exhibit significantly different approaches (based on eye-fixation dwell times, fixation order, interest-area regressions, interest-area eye dwell times) to graph reading.
Interestingly, most study participants exhibited similar eye-track metrics while examining graph after being prompted to find specific information. However, novices and experts show very different eye-track behaviors when they are asked to examine a graph without a specific prompt; the expert behavior remains largely the same as under the prompted conditions, but the novice behavior does not. Analyses of "think-alouds" during the eye-track experiments suggest that experts, with their more developed metacognitive skills, more commonly engage in self-questioning, narrative construction, monitoring, and self-assessment while examining graphs.

Presentation Media

Lindgren & Wirth Poster Presentation (Acrobat (PDF) 20.6MB Jul25 15)