Catalog

Guideline Catalog

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

781 records

Page 29 of 33

  1. Use distinct hues when readers must track intertwined lines

    For trend comparison in dense multi-line displays, use hue-based line colors on ordered-time charts to improve readability and address hard-to-follow shade ramps for readers tracking the same series across crossings or panels.

    • purpose:refine
    • basis:heuristic
    • chart:line
    • time:ordered-time
    • density:dense
    • +3

  2. Use distinctive color hues for categorical data

    For distinguishing unordered groups, use distinctive color hues on categorical color encodings to improve fidelity and mitigate false order cues for readers interpreting category membership.

    • purpose:refine
    • basis:heuristic
    • data:categorical
    • quality:fidelity
    • lever:encoding
    • +1

  3. Use Dorling cartograms instead of rectangular cartograms for big-picture pattern summary

    For overview tasks on geospatial cartograms, prefer a Dorling cartogram type on area-encoded map views to improve insight and mitigate broad-pattern reading errors for readers summarizing spatial trends and distributions.

    • purpose:refine
    • basis:empirical
    • chart:map
    • data:geospatial
    • quality:insight:use
    • +3

  4. Use dot plots instead of bar graphs for average comparisons with unequal group sizes

    For compare tasks on grouped quantitative displays where readers judge averages across groups of different sizes, use dot plots on raw-value charts to improve fidelity and mitigate summed-area judgments for readers interpreting group means.

    • purpose:select
    • basis:empirical
    • task:compare
    • scope:grouped-result
    • chart:dotplot:use
    • +4

  5. Use ensemble-member encoding when uncertainty boundaries could imply physical growth

    For area-level interpretation of geospatial uncertainty over time, use ensemble-member encoding on maps to improve fidelity and mitigate physical-growth misreadings for novice audiences.

    • purpose:refine
    • basis:empirical
    • chart:map
    • data:geospatial
    • time:ordered-time
    • +4

  6. Use equal-scale bars instead of donuts for share comparison

    For compare tasks on record lists of part-to-whole percentages, prefer equal-scale bar charts over donut charts to improve readability and mitigate weak within-record and between-record comparisons for readers scanning across categories and across records.

    • purpose:select
    • basis:heuristic
    • task:compare
    • scope:record-list
    • chart:bar:use
    • +3

  7. Use gender-matched restroom person icons instead of head-outline icons for calibrated risk judgments

    For communicating a single future risk estimate, use gender-matched restroom person icons on icon arrays rather than head-outline person icons to improve fidelity between perceived and displayed risk and address miscalibrated judgments for readers with higher numeracy or graphical literacy.

    • purpose:refine
    • basis:empirical
    • scope:single-result
    • quality:fidelity
    • lever:encoding
    • +2

  8. Use gender-matched restroom person icons instead of rectangular blocks when recall matters

    For communicating a single future risk estimate, use gender-matched restroom person icons on icon arrays rather than rectangular blocks to improve recall accuracy and mitigate overestimation for readers with lower numeracy or graphical literacy.

    • purpose:refine
    • basis:empirical
    • task:retrieve
    • scope:single-result
    • quality:readability
    • +3

  9. Use gray for less important elements and reserve accent colors for highlights

    For charts with highlights or context data, use gray on less important elements to improve visual hierarchy and mitigate competition with highlight colors for readers.

    • purpose:refine
    • basis:heuristic
    • quality:readability
    • lever:encoding
    • polish:hierarchy
    • +1

  10. Use grouped bar charts to foreground legend-category comparisons within x-axis groups

    For open-ended interpretation of grouped three-variable data, use a bar chart instead of a line chart on multivariate quantitative graphs to improve insight into legend-category comparisons within each x-axis group and mitigate unintended x–y interaction reading for viewers identifying the main point.

    • purpose:select
    • basis:empirical
    • task:compare
    • chart:bar:use
    • chart:line:avoid
    • +3

  11. Use high-contrast text colors against the background

    For chart text on screens, use high-contrast foreground and background colors to improve accessibility and mitigate unreadable labels in low-light viewing.

    • purpose:refine
    • basis:heuristic
    • quality:accessibility
    • lever:encoding
    • access:contrast:use
    • +2

  12. Use horizon layering when many time series must fit into very small charts

    For compare tasks on dense ordered-time displays, prefer horizon graph layering on small time-series charts to improve readability and mitigate resolution loss from shrinking standard area plots for readers scanning many trends at once.

    • purpose:refine
    • basis:empirical
    • task:compare
    • time:ordered-time
    • data:temporal
    • +3

  13. Use hue differences for categories without inherent order

    For comparison of unordered categorical values, use hue-based color encoding on charts that color categories to improve readability and mitigate false rank cues for readers distinguishing group identity.

    • purpose:refine
    • basis:heuristic
    • data:categorical
    • quality:readability
    • lever:encoding
    • +1

  14. Use hypothetical outcome plots for ambiguous trend judgments

    For trend judgment in ordered-time displays, use animated hypothetical outcome samples on time-series uncertainty charts to improve fidelity and mitigate misreading of sampling variability for novice public audiences.

    • purpose:refine
    • basis:empirical
    • task:trend
    • time:ordered-time
    • quality:fidelity
    • +4

  15. Use icon arrays that show natural frequencies for low-numeracy risk communication

    For probability communication in health decisions, use natural-frequency encoding on risk displays to improve fidelity and mitigate low-numeracy and denominator-neglect errors for readers with low graph or health literacy.

    • purpose:refine
    • basis:empirical
    • operator:uncertainty
    • quality:fidelity:use
    • lever:encoding
    • +2

  16. Use icon arrays to dampen perceived seriousness relative to numerical frequencies

    For risk framing in non-temporal medical risk displays, use icon arrays on paired baseline-and-treatment views to prevent inflated seriousness and helpfulness judgments and address numerator-focused interpretation in medical decision contexts where false fear or exaggerated benefit claims are a concern.

    • purpose:refine
    • basis:empirical
    • task:compare
    • quality:fidelity:use
    • lever:encoding
    • +1

  17. Use iconic uncertainty symbols when category recognition matters more than speed

    For identifying specific uncertainty types on discrete point-symbol displays, use iconic encoding on the symbol set to improve readability and mitigate weak category-to-symbol matching when readers can understand the uncertainty concept and the visual metaphor.

    • purpose:refine
    • basis:empirical
    • task:retrieve
    • operator:uncertainty
    • lever:encoding
    • +2

  18. Use icons for repeated detail after an explicit summary

    For compact record-list tables, use icon encodings for repeated supporting detail on the table when an explicit row summary is already present to improve aesthetics and mitigate dull or text-heavy detail columns for readers scanning secondary information.

    • purpose:refine
    • basis:heuristic
    • scope:record-list
    • chart:table
    • quality:aesthetics
    • +3

  19. Use independent scales in small-multiple time series when ranges differ and local trends matter

    For trend tasks on ordered-time small multiples, use independent scales on each panel to improve readability and mitigate visually flattened change when series ranges differ widely for analysts inspecting within-series trends.

    • purpose:refine
    • basis:empirical
    • task:trend
    • time:ordered-time
    • chart:area
    • +3

  20. Use inline bars only for the most important numeric columns

    For multi-measure tables with numeric columns, use inline bars on the most important table columns to improve overview and mitigate overly wide quantitative columns for readers who still need exact values.

    • purpose:refine
    • basis:heuristic
    • chart:table
    • data:quantitative
    • quality:insight
    • +3

  21. Use less-saturated colors for racial categories

    For non-temporal comparison of racial or regional groups, prefer less-saturated category colors on categorical palettes to prevent loaded color connotations and mitigate strong red, green, or blue associations for readers of race- or region-focused graphics.

    • purpose:refine
    • basis:heuristic
    • data:categorical
    • quality:trust
    • lever:encoding
    • +3

  22. Use lighter more opaque-looking colors for larger quantities on dark backgrounds when the colormap appears to vary in opacity

    For quantitative comparison on dark-background colormaps, use light-more scale order on color scales that appear to vary in opacity to improve reading speed and mitigate high-end misreading for viewers interpreting magnitude from color.

    • purpose:refine
    • basis:empirical
    • quality:readability:use
    • lever:scale-order
    • channel:opacity:use
    • +1

  23. Use lightness shades of one hue when totals matter more than parts

    For part-whole charts with many categories, use color-lightness variation within one hue on category parts to improve readability and address overly confetti-like palettes for readers who should notice totals before parts.

    • purpose:refine
    • basis:heuristic
    • lever:encoding
    • channel:color-lightness:use
    • operator:part-whole
    • +3

  24. Use line charts instead of horizon graphs when similarity should ignore amplitude and vertical offset changes

    For similarity comparison in ordered-time views, use line charts on time-series displays instead of horizon graphs to improve readability and mitigate slower pattern matching for viewers comparing amplitude- and offset-varying signals.

    • purpose:select
    • basis:empirical
    • task:compare
    • time:ordered-time
    • chart:line:use
    • +3