A Heuristic Approach to Value-Driven Evaluation of Visualizations.

Area of Science:

• Data Visualization
• Cognitive Psychology
• Human-Computer Interaction

Background:

• Interpreting data visualizations requires mapping visual features to concepts.
• Colormaps map color dimensions (e.g., hue, lightness) to data quantities.
• Matching visualizations to user's inferred mappings enhances interpretation.

Purpose of the Study:

• Investigate how background color influences inferred color-quantity mappings in colormaps.
• Resolve conflicting prior research on background color effects.
• Understand factors influencing inferred mappings for effective visualization design.

Main Methods:

• Experimental study examining participant inferences of color-quantity relationships.
• Manipulation of colormap opacity perception and background color.
• Analysis of inferred mappings based on participant responses.

Main Results:

• Inferred color-quantity mappings are influenced by perceived colormap opacity and background color.
• A 'dark-is-more' bias occurs when colormaps lack apparent opacity variation.
• An 'opaque-is-more' bias emerges with increased apparent opacity, potentially overriding the 'dark-is-more' bias on dark backgrounds.

Conclusions:

• Colormap design should consider perceived opacity and background color to align with user inferences.
• To ensure robust and intuitive colormaps, select designs with consistent opacity perception.
• Encoding larger quantities with darker colors is recommended when opacity variation is minimal.