Catalog

Guideline Catalog

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

781 records

Page 32 of 33

  1. Use small multiples instead of one crowded chart when each category needs its own space

    For many-category charts where readers need to inspect each category separately, use small multiples on a multi-view layout instead of a single crowded view to improve readability and address category collisions.

    • purpose:select
    • basis:heuristic
    • structure:small-multiples:use
    • structure:single-view:avoid
    • lever:layout-structure
    • +2

  2. Use small multiples instead of overlay when comparing many time points

    For ordered-time comparison across many time points, use small multiples on repeated views of the same variables to improve insight and mitigate occlusion for analysts comparing change over time.

    • purpose:select
    • basis:empirical
    • task:trend
    • time:ordered-time
    • structure:small-multiples:use
    • +3

  3. Use sparklines to show within-row development over time

    For trend tasks in ordered-time table columns, use sparklines on tables to improve trend recognition and mitigate reliance on only two or three timepoints for readers scanning row-by-row changes.

    • purpose:refine
    • basis:heuristic
    • chart:table
    • task:trend
    • time:ordered-time
    • +2

  4. Use split per-series rows for dispersed cross-series comparisons

    For dispersed cross-series value comparison in ordered time, prefer split per-series layouts on multi-series temporal charts to improve readability and mitigate slow tracing through overlaid series for readers doing quick overview analysis.

    • purpose:select
    • basis:empirical
    • task:compare
    • time:ordered-time
    • structure:multi-view:use
    • +3

  5. Use stacked bar charts instead of stacked area charts for negative correlation judgments

    For association judgments, prefer stacked bar charts over stacked area charts on negatively correlated quantitative series to improve fidelity and address weaker discrimination of correlation strength for people comparing opposite-moving values.

    • purpose:select
    • basis:empirical
    • task:relate
    • chart:bar:use
    • chart:area:avoid
    • +3

  6. Use stacked bar charts instead of stacked line charts for negative correlation judgments

    For association judgments, prefer stacked bar charts over stacked line charts on negatively correlated quantitative series to improve fidelity and address weaker discrimination of correlation strength for people comparing opposite-moving values.

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

  7. Use stacked bars for repeated share breakdowns

    For comparing repeated share breakdowns across groups or time, prefer stacked bar charts over multiple pie or donut charts to improve readability and mitigate hard-to-compare repeated proportions for general audiences.

    • purpose:select
    • basis:heuristic
    • chart:bar:use
    • chart:pie-donut:avoid
    • operator:part-whole
    • +3

  8. Use stacked columns for a few evenly spaced dates when totals matter

    For ordered-time part-to-whole reading across a few evenly spaced dates, use stacked columns instead of an area chart when the total is crucial to improve readability and mitigate harder part comparisons for readers comparing totals and shares together.

    • purpose:select
    • basis:heuristic
    • time:ordered-time
    • chart:bar:use
    • chart:area:avoid
    • +4

  9. Use superimposed value-and-uncertainty maps for integrated lookup

    For exact lookup tasks that require combining value and uncertainty, use superimposed encoding on bivariate maps to improve identification accuracy and mitigate error-prone cross-chart matching for readers interpreting uncertainty together with value.

    • purpose:select
    • basis:empirical
    • task:retrieve
    • structure:single-view:use
    • structure:multi-view:avoid
    • +3

  10. Use superposition instead of small multiples for a small number of time points

    For ordered-time comparison across a small number of time points, use superposition on shared axes to improve fidelity and mitigate imprecise cross-panel comparison for analysts comparing change over time.

    • purpose:select
    • basis:empirical
    • task:trend
    • time:ordered-time
    • structure:single-view:use
    • +3

  11. Use tables instead of line charts for exact value lookup

    For exact value lookup in small static two-dimensional displays, use a table instead of a line chart on tabular data to improve fidelity and speed and mitigate imprecise point reading for readers identifying specific values.

    • purpose:select
    • basis:empirical
    • task:retrieve
    • chart:table:use
    • chart:line:avoid
    • +4

  12. Use text axis labels instead of pictograph labels for brief recall tasks

    For brief recall tasks, prefer text labels on chart axes to improve memory fidelity and address weaker category-to-value encoding for viewers reading short, unambiguous category names.

    • purpose:refine
    • basis:empirical
    • task:retrieve
    • chart:bar
    • quality:fidelity
    • +3

  13. Use the same axis range across consecutive plots of the same variables

    For comparison across repeated views of the same variables, use matching axis ranges on consecutive plots to improve fidelity and mitigate false crossings for readers comparing patterns across views.

    • purpose:refine
    • basis:empirical
    • task:compare
    • structure:multi-view
    • quality:fidelity
    • +3

  14. Use the smallest geographic units available

    For overview mapping of regional patterns, use the smallest available geographic units on a choropleth to improve insight and mitigate hidden within-region variation for readers scanning spatial structure.

    • purpose:refine
    • basis:heuristic
    • chart:choropleth
    • data:geospatial
    • quality:insight
    • +2

  15. Use the title or description as a color key

    For visualizations whose titles or descriptions can name the colored groups, use color-coded headline text on title or description copy to improve readability and mitigate separate legend lookup for readers entering through the surrounding text.

    • purpose:refine
    • basis:heuristic
    • quality:readability:use
    • lever:text-annotation
    • component:title:use
    • +2

  16. Use tooltips only as backup labels for secondary categories

    For interactive multi-category charts with limited label space, use tooltips on secondary categories to improve readability and address label crowding for readers exploring the chart while keeping important categories directly labeled.

    • purpose:refine
    • basis:heuristic
    • lever:interaction-access
    • component:tooltip:use
    • component:label:use
    • +2

  17. Use triplet matching to learn perceptual kernels for encoding design

    For perceptual similarity collection in visualization design, prefer triplet matching on encoding palettes to improve kernel fidelity and mitigate high-variance or underexpressive estimates for designers building perceptual models.

    • purpose:refine
    • basis:empirical
    • quality:fidelity
    • lever:interaction-access
    • communication:workflow
    • +1

  18. Use two closed shapes for trend judgments in mixed scatterplots

    For trend judgments in single-view scatterplots, prefer closed shape encodings on heterogeneous point displays to improve fidelity and mitigate slower or less accurate reading in mixed-symbol views for readers making rapid visual analytic judgments.

    • purpose:refine
    • basis:empirical
    • task:trend
    • chart:scatter
    • quality:fidelity
    • +3

  19. Use uncertainty-revealing rendering when certainty is not warranted

    For uncertainty communication in quantitative displays, use uncertainty-revealing rendering on charts to improve trust and mitigate false objectivity for readers.

    • purpose:refine
    • basis:empirical
    • operator:uncertainty
    • lever:encoding
    • communication:credibility
    • +2

  20. Use uniform perceptual spacing in continuous colormaps

    For equal-interval value reading, prefer uniform perceptual spacing on continuous color scales to improve fidelity and mitigate uneven emphasis for viewers comparing equal steps across the scale.

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

  21. Use uppercase only for short chart text

    For labels and headings in charts, use uppercase only on short text elements to improve readability and mitigate dense, hard-to-scan all-caps text for readers skimming labels.

    • purpose:refine
    • basis:heuristic
    • quality:readability
    • lever:text-annotation
    • polish:hierarchy

  22. Use vertical length encoding for geospatial counts when 2D overlap is a problem

    For compare tasks on geospatial count displays with dense overlap and long-tailed values, use vertical length encoding on anchored map marks to improve fidelity and address overplotting for readers comparing both large and small counts.

    • purpose:refine
    • basis:empirical
    • task:compare
    • data:geospatial
    • quality:fidelity
    • +3

  23. Use vertically stacked small multiples instead of superposition for mean and range comparison

    For set-to-set comparison of mean or range during brief inspection, prefer vertically stacked small-multiple layouts on paired bar charts to improve fidelity and mitigate errors from overlapping sets for brief visual inspection.

    • purpose:select
    • basis:empirical
    • task:compare
    • chart:bar
    • structure:small-multiples:use
    • +3

  24. Use violin plots instead of bar charts with error bars for inferential comparisons

    For comparing grouped means with uncertainty, prefer a violin plot on quantitative summary charts to improve fidelity and mitigate within-the-bar bias for viewers making inferential judgments without deep statistical training.

    • purpose:select
    • basis:empirical
    • task:compare
    • chart:box-violin:use
    • chart:bar:avoid
    • +3