Exploring Causes of the Graph-as-picture Error in Students’ Graph Interpretation: Insights from an Eye-tracking Study with Ninth Graders
摘要
The graph-as-picture error is prevalent when students are asked to match contextual graphs and realistic images. Previous research has not yet sufficiently revealed how and why this error emerges in students’ approaches. This paper presents an exploratory empirical study analyzing students’ approaches to matching a contextual graph and a realistic image in the racing car context, focusing on the characteristics of approaches that lead to the graph-as-picture error and those that do not. Using eye movement data and verbal utterances from 19 ninth graders, we traced students’ approaches. We analyzed the characteristics of students’ approaches and developed a model comprising five characteristic approach components that describe how students connect graph, speed, track features, and realistic images. Applying this model to individual students’ approaches, we found that when the graph-as-picture error did not emerge, students engaged in a three-step process: They (1) extracted relevant information about the speed from the graph, (2) translated what this information means for the real-world situation (track features) and (3) linked this back to the realistic images by establishing the connection between the shape of the realistic image and the real-world situation (track features). In contrast, the graph-as-picture error tended to emerge when students made an error in one of these steps or directly merged the graph and realistic image based on visual similarity. Our findings suggest that the graph-as-picture error should be considered in a differentiated, process-oriented way rather than viewed as a simple error category to facilitate tailored support for students.