A Componential Model of Human Interaction with Graphs. H. Effects of the Distances among Graphical Elements

Based on task analyses of people using graphs, Gillan and Lewis (1992) have developed a model that describes how people interact with graphs. The model proposes that for simple tasks (e.g., comparisons and subtraction) and common graphs (e.g., line, scatter, and bar graphs), graph users apply combinations of five component processes — Searching for indicators, Encoding the value of indicators, performing Arithmetic Operations on the values, making Spatial Comparisons among the indicators, and Responding. The model further suggests that the combination and order of the components that the user applies depends on a user's task and the type of graph. The present research investigated two predictions from the model concerning spatial relations in a graph: (1) that response times to answer comparison questions should be sensitive to varying the distance between two indicators, but not to varying the indicator-to-axis distance, and (2) that response times to answer difference questions should be sensitive to the distance between the indicator and the y-axis, but not to the distance between the indicators. In the experiment, subjects used line and bar graphs to answer comparison and difference questions in which the appropriate distances varied systematically. The results of the research supported both predictions, thereby providing empirical validation of the model. In addition, some aspects of the data were not anticipated by the model, suggesting the need to enhance the componential model.