Overlay explicit outlier markers for interval outlier counts

Overlay explicit outlier marks when viewers must compare how many unusual values appear in each interval. For example, draw broad outlier stripes on top of a smoothed colorfield so unusual values stand out while the underlying field still preserves context.

Outlier counting is easier when unusual values are visually boosted instead of buried inside the full signal. The context field still helps readers keep those outliers tied to their surrounding interval.

Mechanism: The overlay makes unusual values visually salient for numerosity estimation, while the smoothed background preserves the broader time-series context.

Evidence: In the outlier experiment, event striping outperformed all other tested displays, and the paper attributes this advantage to explicitly mapping outliers while retaining a contextual colorfield underneath (Albers et al., 2014).

• User Goal: Identify which interval contains the most outliers.
• Task: Compare outlier frequency across known time blocks.
• Data: Quantitative time series with unusual values that should be highlighted.
• Chart Setting: A colorfield-like overview where both context and outlier emphasis matter.
• Success Criterion: Higher accuracy on outlier-count comparisons.

Break it when: The primary task is comparing interval averages or spread instead of counting outliers. Why: The paper reports that event striping underperformed standard colorfields on all summary tasks except outlier detection.

Sacrifice: You make the display less general-purpose for other summary tasks. Risk: A design tuned for outlier emphasis can distract from reading averages or spread. Mitigation: Reserve the overlay for views whose main job is outlier comparison.

Mistake: Rely on a smoothed colorfield alone for outlier-count comparison. Why it fails: Unusual values are not explicitly boosted, so readers have more difficulty seeing and counting them.

Failure Sign: Reviewers miss isolated unusual values or disagree on which interval contains more of them. Quick Check: Show the same data with and without the outlier overlay, then ask which interval has the most outliers. Stronger Test: Repeat the A/B check on cases where outliers are distributed across several intervals rather than concentrated in one.

• Add a separate visual layer for outliers above the background field.
• Make the outlier marks broad enough to be counted quickly.
• Keep a smoothed background field underneath so the interval context remains visible.