The Proximity Compatibility Principle and Its Applications to Human Perception of Graphic Displays

Abstract

When engineering graphs, it is important that the engineer consider what display format will lead to both the most accurate and most rapid processing of information. The proximity compatibility principle (PCP) states that graphs with high display proximity (ex: line graphs in which lines connect data points) are better for tasks where participants must mentally integrate data points and predict trends, and those with low display proximity (ex: bar graphs, whose data points are unconnected) should be better for tasks that require focused attention on one specific data point. Participants in this experiment viewed either a bar or line graph and answered a corresponding true/false question (requiring either focused attention on a data point or integration across the points). ANOVA results support the PCP, finding a significant interaction between graph and question type for the accuracy of graph interpretation. A follow up experiment examining performance on tasks that required a hybrid of integration and focused attention processing revealed that those questions emphasizing trend integration were better supported by line graphs, whereas bar graphs better supported categorical integration.