Catalog

Guideline Catalog

Browse visualization guideline records with sections, labels, and references.

781 records

Page 18 of 33

  1. Reconstruct a coherent representative subset from dense path ensembles

    For overview reading of uncertain geospatial paths over ordered time, prefer a reconstructed representative subset on dense track displays to improve readability and mitigate clutter and incoherent crossings for viewers interpreting spatial uncertainty.

    • purpose:refine
    • basis:empirical
    • time:ordered-time
    • data:geospatial
    • density:dense
    • +3

  2. Recruit diverse viewers for visualization testing

    For interpretation testing, use a diverse participant sample in visualization review sessions to improve fidelity and address audience-specific interpretation gaps for viewers with mixed ages, education, and visualization experience.

    • purpose:refine
    • basis:rhetorical
    • quality:fidelity
    • lever:interaction-access
    • communication:workflow
    • +2

  3. Reduce and shorten columns before publishing a wide table

    For dense record-list tables in responsive layouts, use column reduction and shortening on tables to improve readability and mitigate cramped horizontal scanning for readers on mobile screens.

    • purpose:refine
    • basis:heuristic
    • chart:table
    • quality:readability
    • lever:layout-structure
    • +2

  4. Reduce data aggregation when showing an observed relationship

    For communicating observational relationships in grouped or binned quantitative displays, avoid high data aggregation on relational charts to improve fidelity and mitigate unwarranted causal interpretations for viewers reading summarized data.

    • purpose:refine
    • basis:empirical
    • task:relate
    • scope:grouped-result
    • quality:fidelity:use
    • +3

  5. Reduce mark quantity when similar marks cannot be grouped

    For target-search tasks in dense views, avoid high mark counts on weakly grouped mark displays to improve search speed and mitigate slow serial scanning for readers locating a single target.

    • purpose:refine
    • basis:empirical
    • quality:readability
    • lever:layout-structure
    • reading-mode:lookup
    • +1

  6. Reduce saturation when applying category colors to large filled marks

    For categorical color encoding on charts with large filled marks, prefer lower-saturation category colors on the filled areas to improve readability and mitigate palettes that become overpowering when reused from lines or points.

    • purpose:refine
    • basis:heuristic
    • data:categorical
    • quality:readability
    • lever:encoding
    • +2

  7. Reduce the number of displayed items in temporal trend views

    For trend reading over ordered time, avoid dense item counts on multidimensional trend views to improve fidelity and mitigate clutter-driven misses for viewers.

    • purpose:refine
    • basis:empirical
    • task:trend
    • time:ordered-time
    • density:dense
    • +3

  8. Reduce the number of distinct categories for simultaneous numerosity estimates

    For simultaneous numerosity comparison across categories, prefer fewer visually distinct groups in one display to improve fidelity and mitigate overload from multiple concurrent count estimates for overview readers.

    • purpose:refine
    • basis:empirical
    • task:compare
    • quality:fidelity
    • lever:encoding
    • +2

  9. Reduce value resolution as uncertainty increases

    For decision tasks on bivariate color maps that combine value and uncertainty, use value-suppressing quantization on uncertain regions to improve uncertainty-aware choices and mitigate overinterpretation of noisy data for decision-makers.

    • purpose:refine
    • basis:empirical
    • task:compare
    • quality:trust:use
    • lever:encoding
    • +3

  10. Reinforce that the horizontal axis increases to the right in connected scatterplots

    For reading paired time series in ordered-time connected scatterplots, use explicit horizontal-axis orientation cues on the x-axis to prevent axis-polarity mistakes and mitigate line-chart carryover for novice readers.

    • purpose:refine
    • basis:empirical
    • time:ordered-time
    • chart:scatter
    • quality:fidelity
    • +3

  11. Remove 3-D depth cues for private inspection of the data

    For private inspection of quantitative findings, avoid added 3-D depth cues on line or bar charts of 2-D data to improve readability and mitigate presentation-oriented embellishment for analysts using the graph themselves.

    • purpose:refine
    • basis:empirical
    • audience:analyst
    • lever:encoding
    • quality:readability
    • +1

  12. Remove 3-D depth cues when the chart must support an immediate decision

    For immediate decision support, avoid added 3-D depth cues on line or bar charts of 2-D quantitative data to improve readability and mitigate memory-oriented embellishment for decision-makers using the figure right now.

    • purpose:refine
    • basis:empirical
    • audience:decision-maker
    • reading-mode:exact
    • lever:encoding
    • +2

  13. Remove a conflicting shape cue when a third class is present

    For grouped comparison tasks with multiclass scatterplots, avoid conflicting shape encoding on point marks when an extra class is also present to prevent accuracy loss and mitigate compounded visual complexity for viewers comparing which group is on average higher.

    • purpose:refine
    • basis:empirical
    • task:compare
    • chart:scatter
    • lever:encoding
    • +3

  14. Remove controls that do not support the chart's task

    For interactive chart reading and task completion, avoid irrelevant controls and overly broad interaction on interactive charts to improve accessibility and mitigate unnecessary options and complexity for readers affected by cognitive load.

    • purpose:refine
    • basis:accessibility
    • quality:accessibility
    • lever:interaction-access
    • needs:cognitive
    • +1

  15. Remove distinct colors when position already separates categories

    For multi-category comparison charts, avoid distinct color hues on marks that are already separated by position or gaps to improve readability and mitigate redundant category coding for readers seeing the chart for the first time.

    • purpose:refine
    • basis:heuristic
    • lever:encoding
    • channel:color-hue:avoid
    • group-cardinality:many
    • +2

  16. Remove expressive imagery when the chart should read as objective

    For explanatory charts in static presentation contexts, avoid expressive visual imagery on charts that should present objective information to prevent unintended value messaging and address the mistake of using persuasive framing when neutrality is required for readers evaluating the data.

    • purpose:refine
    • basis:empirical
    • quality:trust
    • lever:encoding
    • communication:framing
    • +1

  17. Remove gratuitous third dimensions

    For explanatory charts where values must be decoded accurately, avoid decorative depth encoding on the visual representation to prevent perceptual obscuring and address hard-to-judge positions for readers making visual comparisons.

    • purpose:refine
    • basis:empirical
    • quality:fidelity:use
    • lever:encoding

  18. Remove nonessential chart clutter from explanatory charts

    For explanatory charts viewed briefly in presentations or handouts, avoid nonessential visual clutter on a chosen chart to improve perceived professionalism and address distracting default styling for viewers reading quickly.

    • purpose:refine
    • basis:empirical
    • quality:aesthetics
    • lever:encoding
    • polish:declutter
    • +1

  19. Remove nonessential realism from analytic visualizations

    For analytic judgment tasks, avoid highly realistic image styling on maps or scientific visuals to improve fidelity and mitigate distraction and unwarranted credibility for viewers interpreting the display.

    • purpose:refine
    • basis:empirical
    • quality:fidelity:use
    • lever:encoding
    • communication:credibility
    • +1

  20. Remove pictographs that do not encode data

    For brief recall and rapid lookup tasks, avoid superfluous imagery on simple bar charts to prevent distraction and mitigate slower and less accurate value extraction for viewers reading values under time pressure.

    • purpose:refine
    • basis:empirical
    • task:retrieve
    • chart:bar
    • quality:readability
    • +2

  21. Remove purely decorative visual elements

    For explanatory charts aimed at mixed audiences, avoid decorative icons, colors, and stylistic embellishments on an already chosen chart to improve readability and mitigate distracting or confusing overdesign for viewers with varied confidence.

    • purpose:refine
    • basis:rhetorical
    • quality:readability
    • lever:encoding
    • polish:declutter
    • +1

  22. Remove redundant place labels when annotations already orient the reader

    For explaining multiple annotated patterns on a detailed map, avoid redundant place labels on the map to improve readability and mitigate annotation clutter for readers who can orient from the annotation text.

    • purpose:refine
    • basis:heuristic
    • chart:map
    • quality:readability
    • lever:text-annotation
    • +2

  23. Remove required interactions from the first reading of a chart

    For initial reading and interpretation, avoid required interaction access on chosen charts to improve fidelity and mitigate extra task demands for novice readers.

    • purpose:refine
    • basis:empirical
    • lever:interaction-access
    • quality:fidelity
    • literacy:novice

  24. Remove single-threshold task labels from slope charts

    For grouped-result comparison of search-task slopes, avoid threshold annotations on slope charts to prevent false task classification and mitigate slope-only misinterpretation for domain experts.

    • purpose:refine
    • basis:empirical
    • task:compare
    • scope:grouped-result
    • quality:fidelity
    • +3