This study empirically examines three display design features that can hinder or promote viewers’ perceptual bias of bar charts. Bar charts are commonly used in various landscapes, including business contexts, yet, research on their impact on bias is limited. In an experiment, 266 undergraduates viewed vertical bar charts with two styles of scaling (fine vs coarse), two formats of bar ordering (in a certain order vs in random order) and two patterns of bar width (same vs different). A 7-item quiz measured the level of perceptual bias. Results showed that the main effects of all three display features. The results revealed that the interaction of bar ordering formats and (1) the scaling styles and (2) the bar thickness patterns on viewers’ perceptual bias were significant. The other effects were not. In addition to extending the theoretical insight into visual deception in the contexts of bar charts, the findings provide practical guidelines for visual makers on bar chart designs that could lead to readers’ perceptual bias, and for visual viewers on how to properly decode the meaning embedded in bar charts.
Read full abstract