The Role of Text in Visualizations: How Annotations Shape Perceptions of Bias and Influence Predictions

Area of Science:

• Human-Computer Interaction
• Information Visualization
• Cognitive Psychology

Background:

• Textual annotations in data visualizations can influence reader interpretation.
• Understanding the impact of semantic content and wording bias is crucial for effective data communication.

Purpose of the Study:

• To investigate how text position, semantic content, and biased wording in visualizations affect data trend prediction and perceived author bias.
• To explore the relationship between textual bias and perceived author bias.
• To develop a crowdsourced method for generating biased chart annotations.

Main Methods:

• Conducted two empirical studies using bar and line charts.
• Employed tasks involving data trend prediction and bias appraisal.
• Utilized a crowdsourced approach to create annotations with varying degrees of bias.

Main Results:

• Textual additions minimally affected data trend perception.
• Significant impact observed on the perception of author bias, correlating with the degree of textual bias.
• Exploratory analyses indicated an interaction between data prediction and perceived bias.

Conclusions:

• Authorial bias expressed through text in visualizations strongly influences reader perception of that bias.
• Designers must be mindful of textual elements to avoid polarizing reader opinions.
• The developed crowdsourced method offers a tool for creating controlled biased annotations for research.