Ranking Visualizations of Correlation Using Weber's Law.

Area of Science:

• Human-Computer Interaction
• Cognitive Psychology
• Data Visualization

Background:

• Effective data visualization design remains challenging for practitioners.
• Leveraging perceptual laws offers a quantitative approach to evaluate visualization effectiveness.

Purpose of the Study:

• To investigate if Weber's law can model correlation perception in common visualizations.
• To quantitatively assess and compare the perceptual precision of different visualization designs.

Main Methods:

• A large-scale crowdsourced experiment with 1687 participants.
• Investigated perception of correlation in nine common visualization types.
• Applied Weber's law to model the precision of correlation judgments.

Main Results:

• Precision of correlation judgment was successfully modeled by Weber's law for all tested visualizations.
• Significant variation in correlation judgment precision was observed between negatively and positively correlated data.
• Weber models provide a quantitative method to rank visualization perceptual precision.

Conclusions:

• Weber's law is applicable for modeling perceptual precision in data visualization.
• Understanding variations in judgment precision is crucial for effective visualization design.
• Quantitative evaluation using perceptual laws enhances visualization selection and design.