An exploratory study on eye-gaze patterns of experts and novices of science inference graph items

Abstract

<abstract><p>Graphs are highly prevalent as a form of quantitative data in various science, technology, engineering and mathematics fields. Thus, graphical literacy is especially important in understanding today's world and being scientifically literate. However, students often face difficulties in graph interpretation and differ substantially in their graphical literacy. While many teachers are aware of students' difficulties in answering graph items, there is limited knowledge about how students go about attempting graph items. In this exploratory study, we investigated the eye-gaze patterns of experts and novices in graph interpretation of five science inference-based multiple-choice items requiring no prior content knowledge to problem-solve. Experts refer to science university faculty members who are currently teaching science content courses to undergraduate students. Novices refer to university undergraduates majoring in one of the science subjects. Participants' eye-gaze movements were recorded using the Dikablis eye-tracker, and their eye-gaze patterns and total glance time (s) were subsequently analyzed using the software D-Lab 3.0. Experts focused more on the question stem, whereas novices focused more on the graph. Additionally, experts tend to focus on contextual and graph data features initially, before moving to cues such as options. Conversely, novices demonstrated more sporadic search patterns. The findings contribute to the literature that compares how experts and novices' problem-solve graph items that are inference-based. An interesting future study on the eye gaze patterns and accuracy of answers is suggested from a finding. This study also provides a set of heuristics to be adopted in the teaching and learning of graph interpretation. The findings of this study have implications for teachers in the way they scaffold students' approach to answering graphical items. Additionally, students can employ heuristics to answer graphical items more effectively.</p></abstract>