A Critical Reflection on the Values and Assumptions in Data Visualization

Paper Title

A Critical Reflection on the Values and Assumptions in Data Visualization

Publication Info

• Topic area: Examination of foundational values in data visualization research and practice.
• Keywords: data visualization, universality, objectivity, efficiency, normative visualization, critical reflection, design values, visualization tools, pedagogy, interdisciplinary critique.

Background and Problem

• Problem / challenge: The field of data visualization has been shaped by values such as universality, objectivity, and efficiency, which are deeply embedded in its tools, guidelines, and research practices. However, these values may limit the field's ability to address diverse perspectives and contexts.
• Significance: Understanding and critiquing these values is crucial for evolving visualization practices to better serve diverse audiences and contexts, fostering innovation and inclusivity.
• Motivation and related work: Foundational texts by Bertin, Tukey, Wilkinson, Ware, and Munzner have established the dominant paradigm of "normative visualization," emphasizing scientific and engineering values. However, critiques from fields like data feminism, humanistic visualization, and cultural studies highlight the need for alternative priorities and approaches.

Solution

• Proposed approach: A critical reflection on the values of universality, objectivity, and efficiency as articulated in five seminal texts in data visualization, with a call for a broader, pluralistic value system.
• Novelty:
  1. Identification and articulation of three dominant values (universality, objectivity, efficiency) in normative visualization.
  2. Examination of how these values shape visualization tools, guidelines, and practices.
  3. Integration of critiques from interdisciplinary perspectives to highlight the limitations of these values.
  4. Proposal for expanding the value landscape of visualization research and practice.
• Procedure and key techniques:
  • Selection of five seminal texts: Bertin’s Semiology of Graphics, Tukey’s Exploratory Data Analysis, Wilkinson’s The Grammar of Graphics, Ware’s Visual Thinking for Design, and Munzner’s Visualization Analysis and Design.
  • Analysis of prefaces, introductions, and first chapters to identify value-laden statements.
  • Collaborative, iterative discussions among authors with diverse academic backgrounds.
  • Use of diffractive reading and synthesis to trace the influence of these values on contemporary visualization practices.

Results

• Concrete findings:
  • Universality: Visualization assumes a shared perceptual baseline, aiming for universally readable designs. However, critiques highlight cultural and cognitive differences that challenge this assumption.
  • Objectivity: Visualization is framed as a neutral tool for truth discovery, but critiques argue that data and visualizations are inherently situated and interpretive.
  • Efficiency: Visualization prioritizes rapid and accurate insights, yet alternative approaches value slower, more reflective, and affective engagements.
• Advantage over baselines: The paper does not propose a new system but critiques the dominant paradigm, offering a pathway to more inclusive and context-sensitive visualization practices.
• Experiments / evaluation: The analysis is qualitative, based on textual interpretation and collaborative discussion rather than empirical experiments.
• Limitations and future work:
  • The analysis is limited to five seminal texts and reflects the authors' positionalities and disciplinary backgrounds.
  • The identified values are not exhaustive but represent a recurring configuration in normative visualization.
  • Future work should explore additional values and their implications for visualization research and practice.

Summary

This paper critically examines the values of universality, objectivity, and efficiency in normative data visualization, as articulated in five seminal texts. These values have guided decades of innovation but face critiques for their limitations in addressing cultural, cognitive, and contextual diversity. The authors call for an expanded value landscape that includes subjectivity, individuality, and alternative priorities, fostering more inclusive, imaginative, and context-sensitive visualization practices. This reflection aims to enrich pedagogy, tools, and research by integrating diverse perspectives and challenging the dominance of normative paradigms.