Abstract
When people interpret colormap data visualizations, they are influenced by expectations of how colors map to quantities. These expectations are constructed from multiple biases, including the dark-is-more bias (darker colors represent larger quantities) and the opaque-is-more bias (regions appearing more opaque represent larger quantities). However, past studies of the opaque-is-more bias tested colormaps that varied in lightness (Schloss et al. 2019), so it is unknown whether the opaque-is-more bias operates in absence of lightness variation. To address this question, we presented participants with colormap data visualizations that varied in apparent opacity, while holding lightness constant (L*=50). Colormaps ranged from gray to a high chroma color (blue or green) and were presented on a gray or chromatic background matching the hue in the colormap (e.g., gray-blue map on blue background). Thus, higher chroma colors appeared more opaque on the gray background, and lower chroma colors appeared more opaque on the chromatic background. The side of the map appearing more opaque was left/right balanced. Participants were told the colormaps represented measurements in different counties, and were asked to indicate whether measurements were larger on the left/right of the map. Participants were more likely than chance to choose the more opaque side, providing evidence for the opaque-is-more bias in absence of lightness variation. However, this likelihood was greater when the more opaque region had higher chroma, suggesting a possible “high chroma-is-more bias.” To test this possibility, we removed opacity variation by presenting the same colormaps on alternate chromatic backgrounds (e.g., gray-blue map on green background). Participants selected the chromatic side significantly more than chance, providing evidence for a new, high chroma-is-more bias in absence of lightness and opacity variation. These findings extend our knowledge of how people infer meaning from visual features, and can be used to inform design of effective information visualizations.