Data Visualization

emerging Updated: 2026-05-31
data-visualization graphical-perception pre-attentive-processing gestalt-principles color color-deficiency tufte data-ink chartjunk storytelling narrative interaction big-data human-computer-interaction

Data Visualization

Status: emerging
Last updated: 2026-05-31
Sources: 9781119636113.Ch35.Pdf
Tags: [data-visualization, graphical-perception, pre-attentive-processing, gestalt-principles, color, color-deficiency, tufte, data-ink, chartjunk, storytelling, narrative, interaction, big-data, human-computer-interaction]

Summary

Data visualization is the graphical representation of data that communicates the information in it to human observers, exploiting the brain's greater facility with visual information than with numbers or words (Pattanaik & Wiegand, 2021). The chapter treats it as both an art and a science: it sets out the rules and best practices for building effective visualizations and, where possible, grounds them in the workings of the human visual system. It moves from objectives (audience, purpose, narrative) through principles (graphical perception, design, color, and Tufte's principles of maximizing data), techniques for different data types, interaction, and tools, to future directions. The recurring argument is that a visualization is a deliberate abstraction built to serve a specific point in a larger narrative, not a default output of a plotting tool.

Body

Context

Pattanaik and Wiegand (2021), in chapter 35 of the Handbook of Human Factors and Ergonomics (Part 7, Human–Computer Interaction), examine how to construct effective data visualizations and why particular choices work, reasoning from the limits and strengths of the human visual system. The chapter is written for researchers, students, and practitioners in human factors who must disseminate data, and it spans theory (graphical perception, color, Tufte's principles), a catalogue of plot types, interaction, and tooling. Within this knowledge base the article sits in the interface-and-interaction domain: it is the perceptual and aesthetic counterpart to Representation Design, which treats display design for meaning processing; it operationalises the perceptual mechanisms described in Information Processing and Sensation And Perception for the specific task of plotting data; and its treatment of grouping draws directly on the Gestalt principles formalised in Perceptual Organization.

Key Points

Visualization is purpose-driven abstraction. The first and most important principle is to know the audience and the purpose of the visualization; good plots play a clear role in a larger narrative rather than dumping data on the reader (PDF p. 2, orig. p. 896). Producing a high-quality visualization should resemble writing a paper — consider audience, develop a clear objective, identify the visual encodings that serve the narrative, then iterate — rather than the common shortcut of selecting data and accepting a tool's default template (PDF p. 2, orig. p. 896). Visualization is by nature a reduction of information: when data are visualized, data are lost, so the designer must consider both what the reader should see and what becomes invisible, and treat the omission as an ethical as well as a design choice (Munzner, 2014, as cited in Pattanaik & Wiegand, 2021) (PDF p. 4, orig. p. 898). Context and annotation can draw attention and reveal subtleties, with the trade-off of clutter, so annotation is used sparingly (PDF p. 4, orig. p. 898). The chapter situates this in a long history of graphical communication, crediting William Playfair as the most influential figure for the line, bar, area, and pie charts (Berkowitz, 2018, as cited in Pattanaik & Wiegand, 2021), and notes the narrative emphasis advanced by writers such as Yau (2011) and Nussbaumer Knafic (2015) (PDF pp. 1–4, orig. pp. 895–898).

Graphical perception sets the limits. The human visual system detects a limited set of visual properties pre-attentively — before conscious awareness — including unique single features, boundaries between groups, and motion of uniquely marked regions; once features are compounded (a conjunction of shape and color), pre-attentive search breaks down (Healey & Enns, 2012, as cited in Pattanaik & Wiegand, 2021) (PDF p. 7, orig. p. 901). Sustained or repeated visual search makes the observer less efficient and prone to false patterns, illustrated by the Where's Waldo books, so the designer should not make readers search for what they need but draw attention to it (Wolfe, 1994; Handford, 1987, as cited in Pattanaik & Wiegand, 2021) (PDF p. 7, orig. p. 901). For encoding numbers, Cleveland and McGill (1984) give a precision hierarchy in which position along a common scale is decoded most accurately, ahead of length, angle, area, and finally color (as cited in Pattanaik & Wiegand, 2021) (PDF p. 8, orig. p. 902). Few (2009) complements this by organising pre-attentive visual attributes into categories — form, color, spatial position, and motion — so designers can see which channels carry categorical versus quantitative distinctions (as cited in Pattanaik & Wiegand, 2021) (PDF p. 8, orig. p. 902).

Design principles and Gestalt grouping. General design guidance is framed around questions of affordance, aesthetics, and learnability, with the aesthetic-usability effect noted (Lidwell, Holden & Butler, 2003; Kurosu & Kashimura, 1995, as cited in Pattanaik & Wiegand, 2021) (PDF p. 12, orig. p. 906). Gestalt grouping principles — proximity, similarity, connectedness, continuity, closure — let the designer create visual grouping metaphors so that separate marks are read as wholes; connectedness and continuity can override proximity and similarity, and connectedness can also create problems in dense networks with overlapping edges (Wong, 2010a, 2010b, as cited in Pattanaik & Wiegand, 2021) (PDF p. 12, orig. p. 906).

Color. Color encodes numeric quantity, categorical distinction, and aesthetic appeal. People generally prefer cool colors (blue most preferred, dark yellow least), and warm colors carry associations of importance and salience, so they are used selectively to advance data of interest rather than as the main display color (Palmer & Schloss, 2010, as cited in Pattanaik & Wiegand, 2021) (PDF p. 14, orig. p. 908). Standard color names are far too few — the visual system distinguishes on the order of a million colors — so digital 3D color models such as RGB are used for representation and transmission (Gibson et al., 2017, as cited in Pattanaik & Wiegand, 2021) (PDF p. 14, orig. p. 908). About 8% of the male population has some form of color deficiency, predominantly red–green, so using color-blind-friendly palettes is a matter of accessibility (ColourBlindAwareness.org, n.d., as cited in Pattanaik & Wiegand, 2021) (PDF p. 18, orig. p. 912).

Tufte's principles of maximizing data. The chapter presents Tufte's principles for putting data first (Tufte, 1990, 2001, as cited in Pattanaik & Wiegand, 2021) (PDF p. 22, orig. p. 916). Graphical integrity requires that the representation of numbers be proportional to the quantities; Tufte's lie factor (size of effect shown in the graph divided by size of effect in the data) should equal 1, and common distortions arise from 3D perspective foreshortening, improper area/volume mapping, and truncating zero from an axis. The data-ink ratio should approach 1, achieved by erasing non-data ink such as heavy gridlines and backgrounds. Chartjunk — decorative embellishment — should be minimised, data density maximised, and small multiples used to show many slices of a dataset compactly. The chapter qualifies the anti-embellishment stance with Bateman et al. (2010), whose study found embellished charts can aid recall, while noting that effective embellishment demands rare artistic skill (as cited in Pattanaik & Wiegand, 2021) (PDF p. 22, orig. p. 916).

Techniques, interaction, and tools. The chapter catalogues techniques by data type — 1D (strip and point plots with jitter and transparency to manage overlap), 2D, multidimensional, volume, vector-field, network, molecule, and uncertainty visualization, including quantile dot plots (PDF pp. 24–40, orig. pp. 918–934). Interaction lets the viewer explore and manipulate the data, organised around Shneiderman's visual-information-seeking mantra, "overview first, zoom and filter, details on demand" (Shneiderman et al., 2016, as cited in Pattanaik & Wiegand, 2021), with interaction itself treated as a research topic (Dimara & Perin, 2020; Hornbæk & Oulasvirta, 2017, as cited in Pattanaik & Wiegand, 2021) (PDF pp. 40–42, orig. pp. 934–936). Future directions point to Big Data handling and visualization and to non-traditional disciplines such as the digital humanities, including character-network analysis of narratives (Cutting, 2016; Helic et al., 2011, as cited in Pattanaik & Wiegand, 2021) (PDF p. 49, orig. p. 943).

Conclusion

Pattanaik and Wiegand (2021) conclude that data visualization is a transdisciplinary practice combining aesthetics, psychology, computer science, and statistics, and that it is the primary method for finding meaning in data yet is often treated with less care than the writing it accompanies. Effective and clear visualizations are integral parts of a narrative, and distilling complex data into pictures that convey meaning is as serious an endeavour as the writing itself. With increasingly powerful tooling and more data available, the chapter's takeaway is that the discipline of audience, purpose, perceptual principle, and honest abstraction — not the tool — determines whether a visualization communicates (PDF p. 50, orig. p. 944).

References

Bateman, S., Mandryk, R., Gutwin, C., Genest, A., McDine, D. & Brooks, C. (2010) 'Useful junk? The effects of visual embellishment on comprehension and memorability of charts', in Proceedings of the SIGCHI Conference on Human Factors in Computing Systems. To be validated.

Berkowitz, B. (2018) Playfair: The True Story of the British Secret Agent Who Changed How We See the World. Washington, DC: George Mason University. To be validated.

Cleveland, W. & McGill, R. (1984) 'Graphical perception: Theory, experimentation, and application to the development of graphical methods', Journal of the American Statistical Association, 79(387), pp. 531–554. To be validated.

ColourBlindAwareness.org (n.d.) Colour blindness. Available at: http://www.colourblindawareness.org/colour-blindness/. To be validated.

Cutting, J. (2016) 'Narrative theory and the dynamics of popular movies', Psychonomic Bulletin & Review, 23(6), pp. 1713–1743. To be validated.

Dimara, E. & Perin, C. (2020) 'What is interaction for data visualization?', IEEE Transactions on Visualization and Computer Graphics. To be validated.

Few, S. (2009) Now You See It: Simple Visualization Techniques for Quantitative Analysis. 1st edn. Oakland, CA: Analytics Press. To be validated.

Gibson, E., Futrell, R., Jara-Ettinger, J., Mahowald, K., Bergen, L., Ratnasingam, S. et al. (2017) 'Color naming across languages reflects color use', PNAS, 114(40), pp. 10785–10790. To be validated.

Handford, M. (1987) Where's Waldo? Boston: Little, Brown and Company. To be validated.

Healey, C. & Enns, J. (2012) 'Attention and visual memory in visualization and computer graphics', IEEE Transactions on Visualization and Computer Graphics, 18(7), pp. 1170–1188. To be validated.

Helic, D., Trattner, C., Strohmaier, M. & Andrews, K. (2011) 'Are tag clouds useful for navigation? A network-theoretic analysis', International Journal of Social Computing and Cyber-Physical Systems, 1. To be validated.

Hornbæk, K. & Oulasvirta, A. (2017) 'What is interaction?', in CHI '17: Proceedings of the 2017 CHI Conference on Human Factors in Computing Systems, pp. 5040–5052. To be validated.

Kurosu, M. & Kashimura, K. (1995) 'Apparent usability vs. inherent usability: Experimental analysis on the determinants of the apparent usability', paper presented at CHI '95. To be validated.

Lidwell, W., Holden, K. & Butler, J. (2003) Universal Principles of Design. New York: Rockport Publishers. To be validated.

Munzner, T. (2014) Visualization Analysis and Design. New York: O'Reilly. To be validated.

Nussbaumer Knafic, C. (2015) Storytelling with Data. Hoboken, NJ: Wiley. To be validated.

Palmer, S. & Schloss, K. (2010) 'An ecological valence theory of human color preference', Proceedings of the National Academy of Sciences of the United States of America, 107(19), pp. 8877–8882. To be validated.

Pattanaik, S.N. & Wiegand, R.P. (2021) 'Data Visualization', in Salvendy, G. & Karwowski, W. (eds.) Handbook of Human Factors and Ergonomics. 5th edn. Hoboken, NJ: John Wiley & Sons, ch. 35, pp. 895–946. doi: 10.1002/9781119636113.ch35. pattanaik2021datavisualization

Shneiderman, B., Plaisant, C., Cohen, M., Jacobs, S., Elmqvist, N. & Diakopoulos, N. (2016) Designing the User Interface: Strategies for Effective Human-Computer Interaction. 6th edn. Harlow: Pearson. To be validated.

Tufte, E. (1990) Envisioning Information. New York: Graphics Press. To be validated.

Tufte, E. (2001) The Visual Display of Quantitative Information. New York: Graphics Press. To be validated.

Wolfe, J. (1994) 'Guided search 2.0: A revised model of visual search', Psychonomic Bulletin & Review, 1(2), pp. 202–238. To be validated.

Wong, B. (2010a) 'Gestalt principles (Part 1)', Nature Methods, 7, p. 683. To be validated.

Wong, B. (2010b) 'Gestalt principles (Part 2)', Nature Methods, 7, p. 941. To be validated.

Yau, N. (2011) Visualize This. Hoboken, NJ: Wiley. To be validated.

Open Questions

  • The chapter presents Tufte's anti-embellishment principles alongside Bateman et al. (2010), whose evidence that embellishment can aid memorability cuts the other way. When does chartjunk help rather than distract, and how should the designer weigh recall against decoding accuracy?
  • Cleveland and McGill's (1984) encoding hierarchy and the pre-attentive accounts (Healey & Enns, 2012; Wolfe, 1994) come from separate research traditions. How well do the precision hierarchy and the pre-attentive feature set agree on which channels to use for a given task?
  • The chapter treats interaction largely descriptively (Shneiderman's mantra, tool surveys). What measurable effect does interactivity have on comprehension versus exploration, an open question the cited interaction literature (Dimara & Perin, 2020; Hornbæk & Oulasvirta, 2017) itself raises?