Using Counterfactuals to Improve Causal Inferences From Visualizations.

Area of Science:

• Data Visualization
• Causal Inference
• Human-Computer Interaction

Background:

• Traditional data visualization focuses on comparison and exploration, aiding correlation identification.
• Users often incorrectly infer causal relationships from visualizations.
• This necessitates methods for direct visual causal inference.

Purpose of the Study:

• To review recent advances in visual causal inference methods.
• To identify limitations of current approaches.
• To outline open challenges and research priorities in the field.

Main Methods:

• Review of recent research in visual causal inference.
• Analysis of limitations in existing visual causal inference techniques.
• Identification of key open challenges for future research.

Main Results:

• Recent research has developed methods to directly support visual causal inference.
• Current methods have limitations restricting their real-world applicability.
• Several key open challenges remain in advancing visual causal inference.

Conclusions:

• Visual causal inference is an emerging area with potential to improve data interpretation.
• Further research is needed to overcome limitations and develop robust methods.
• Addressing open challenges will advance the state of the art in visual causal inference.