Data Storytelling in Data Visualisation: Does it Enhance the Efficiency and Effectiveness of Information Retrieval and Insights Comprehension?

Title of the Paper

Data Storytelling in Data Visualization: Does it Enhance the Efficiency and Effectiveness of Information Retrieval and Insights Comprehension?

Paper Information

• Subject Area: Data Visualization and Data Storytelling
• Keywords: Data Storytelling, Data Visualization, Data Comprehension, Visualization Literacy, User Study, Information Retrieval, Data Cognition

Research Background and Issues

• Identified Problems or Challenges:

  1. Data Storytelling (DS) combines visual elements with narrative to enhance the understanding of complex data for general users. However, its advantages in improving efficiency and effectiveness lack comprehensive empirical support.
  2. The general public often lacks Visualization Literacy, making it challenging to interpret data visualizations. This issue positions data storytelling as a potential tool to enhance data comprehension.
  3. Research on the potential benefits of data storytelling shows mixed results, such as its effects on memory recall, empathy, and comprehension tasks.

• Significance of the Research: In a data-driven world, efficiently and effectively communicating critical data information is essential for decision-making and public education. Understanding whether data storytelling can comprehensively improve efficiency and accuracy, and how it impacts users with varying levels of literacy, provides a theoretical foundation for designing better data visualizations.

• Research Motivation and Related Work:

  • Motivation: To address the current research gap regarding the impact of data storytelling on efficiency and effectiveness.
  • Related Work: Previous studies suggest that data storytelling can convey information more intuitively (e.g., explaining complex topics, enhancing attention to trends). However, these studies are often theoretical or preliminary, lacking systematic empirical validation.

Proposed Solution

• Proposed Solution: The authors designed and conducted a comparative experiment with 103 participants to systematically validate the specific performance of data storytelling in terms of efficiency and effectiveness.

• Innovative Aspects:

  1. Examining the efficiency and effectiveness of data storytelling at two levels: information retrieval (identifying specific data points) and data comprehension (insights generation).
  2. Quantitatively analyzing the potential influence of individual visualization literacy on this relationship.
  3. Conducting the first comprehensive validation of user perception and the role of specific design elements in data storytelling (e.g., color emphasis, annotations, titles).

• Implementation Steps and Key Techniques:

  1. Experiment Design: Comparing six pairs of conventional data visualizations with their corresponding data storytelling versions. Tasks included information retrieval and single/multiple insights comprehension, with both qualitative and quantitative data collection.
  2. Data Usage: Content was selected from publicly available datasets on global climate change, and standard software (Tableau and LucidChart) was used to create dynamic and static data visualizations.
  3. Measurement Metrics:
     • Efficiency: Quantified by task completion time.
     • Effectiveness: Evaluated by correct response rate (Correct Rate, adjusted for guessing factors).
  4. Analysis Methods: Wilcoxon signed-rank test, regression analysis, and other methods were employed to explore data connections.

Research Findings

• Specific Findings:

  1. Data storytelling significantly improved task effectiveness (accuracy) but did not significantly reduce task completion time.
  2. Data storytelling demonstrated higher efficiency in single-insight tasks, making it suitable for simple and manageable visualization scenarios.
  3. Certain data storytelling elements received positive feedback from some users (e.g., enhancing recognition of target data points), while others found them overly complex and distracting.
  4. Visualization literacy significantly influenced users' task performance (accuracy), but no significant interaction was observed between literacy levels and the presence of data storytelling.

• Advantages Compared to Existing Solutions: This study is the first to systematically validate whether data storytelling can substantively enhance users' retrieval and insight capabilities, beyond studies focusing solely on memory or engagement.

• Experimental or Evaluation Results:

  1. The correct rate of the data storytelling version (median: 0.889, IQR: 0.333) was significantly higher than that of the conventional visualization version (median: 0.667, IQR: 0.334).
  2. Efficiency in information retrieval and single-insight tasks showed slight improvement, but no significant differences were observed in multiple-insight tasks.

• Limitations and Future Directions:

  1. Limitations:
     • The experiment was limited to static charts and did not include interactive data storytelling.
     • The dataset did not balance participants' global backgrounds.
     • The range of chart types and task complexity was limited.
  2. Future Directions:
     • Investigate more complex chart types and visualization tools.
     • Introduce interactive elements to explore richer scenarios.
     • Incorporate language and cultural variables to analyze how data storytelling design adapts to users from different cultural backgrounds.
     • Explore the relationship between data storytelling and unresolved areas such as memory retention and empathy.